regulatory impact assessment for proposed …regulatory impact assessment for proposed hazardous...

REGULATORY IMPACT ASSESSMENTFOR PROPOSED HAZARDOUS WASTE

COMBUSTION MACT STANDARDS

DRAFT

Prepared for:

Office of Solid WasteU.S. Environmental Protection Agency

401 M Street, SWWashington, DC 20460

Prepared by:

Industrial Economics, Incorporated2067 Massachusetts Avenue

Cambridge, MA 02140

November 13, 1995

TABLE OF CONTENTS

EXECUTIVE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ES-1

ES1.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ES-1ES2.1 Summary of Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ES-1ES3.1 Economic Impact Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ES-5

ES3.1.1 Economic Impact Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . ES-6ES3.1.2 Differential Impacts of the Rule by Market Sector . . . . . . . . . . . . ES-8ES3.1.3 Uncertainties With Economic Impact Assessment . . . . . . . . . . . ES-13

ES4.1 Benefits of the Rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ES-13

ES4.1.1 Cost Effectiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ES-13ES4.1.2 Human Health Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ES-14ES4.1.3 Ecological Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ES-14ES4.1.4 Property Value Benefits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ES-15

ES5.1 Regulatory Flexibility and Environmental Justice . . . . . . . . . . . . . . . . . . . . . . . . ES-15

ES6.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ES-16

INTRODUCTION AND REGULATORY OPTIONS . . . . . . . . . . . . . . . . . . . . . . . . . CHAPTER 1

1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-11.2 Regulatory Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-21.3 Organization of this Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-6

OVERVIEW OF CURRENT COMBUSTIONPRACTICES AND MARKETS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CHAPTER 2

2.1 Overview of Services Provided by the Hazardous Waste Combustion Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1

2.1.1 Regulatory Requirements Encouraging Combustion . . . . . . . . . . . . . . . . . . 2-22.1.2 Liability Concerns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-32.1.3 Economic Forces Encouraging Combustion . . . . . . . . . . . . . . . . . . . . . . . . 2-3

2.2 Hazardous Waste Combustion Facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-42.3 The Quantity and Characteristics of

Combusted Hazardous Waste . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-72.4 Major Sources of Combusted Wastes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-92.5 Baseline Emissions and Emission Controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-9

2.5.1 Baseline Emissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-92.5.2 Baseline Pollution Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-12

TABLE OF CONTENTS(Continued)

2.6 Combustion Market Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-12

2.6.1 Overcapacity and Effects on Poor Market Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-14

2.6.2 Existing Market Advantages for BIFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-162.6.3 Market Performance Across Combustion Sectors . . . . . . . . . . . . . . . . . . . 2-182.6.4 Financial Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-192.6.5 Capacity Expansions and Cancellations . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-21

2.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-22

ENGINEERING COST ANALYSIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CHAPTER 3

3.1 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-1

3.1.1 Model Plants Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-13.1.2 Compilation of Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-10

3.2 Results of Model Plants Cost Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-12

3.2.1 Compliance Costs-Engineering Estimates . . . . . . . . . . . . . . . . . . . . . . . . . 3-123.2.2 Average Costs Per Combustion Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-183.2.3 Cost Reductions Associated with Feed Reduction . . . . . . . . . . . . . . . . . . . 3-183.2.4 MACT Costs for New Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-18

3.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-21

ECONOMIC IMPACT ANALYSIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CHAPTER 4

4.1 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-2

4.1.1 Data Inputs for Economic Impact Analysis . . . . . . . . . . . . . . . . . . . . . . . . . 4-34.1.2 Method for Analyzing Waste Management Alternatives . . . . . . . . . . . . . . . 4-94.1.3 Method for Analyzing the

Impact of Waste Minimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-134.1.4 Economic Impact Screens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-144.1.5 Break Even Quantity Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-174.1.6 Method for Analyzing the Impact of Feed Reductions . . . . . . . . . . . . . . . 4-254.1.7 Method for Analyzing Removal of

Small Quantity Burner Exemption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-26

4.2 Results of Economic Impact Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-27

4.2.1 Results of Basic Economic Screens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-274.2.2 Alternatives to Combustion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-31


4.2.3 Break Even Quantity Analysis and Changes inWaste Burning Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-42

4.2.4 Social Costs of the Proposed MACT Standards . . . . . . . . . . . . . . . . . . . . . 4-654.2.5 Waste Feed Reduction Compliance Option . . . . . . . . . . . . . . . . . . . . . . . . 4-664.2.6 Effects of Removing Small

Quantity Burner Exemption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-73

4.3 Summary of Economic Impacts on Affected Groups . . . . . . . . . . . . . . . . . . . . . . . 4-73

4.3.1 Key Conclusions Applicable toEntire Combustion Market . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-76

4.3.2 Differential Impacts of the Rule by Market Sector . . . . . . . . . . . . . . . . . . 4-774.3.3 MACT in Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-82

COST EFFECTIVENESS AND BENEFITS ANALYSIS . . . . . . . . . . . . . . . . . . . . . . CHAPTER 5

5.1 Cost Effectiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1

5.1.1 Cost Effectiveness Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-15.1.2 Cost Effectiveness Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-4

5.2 Health Benefits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-9

5.2.1 Exposure to Dioxin Emissions and Health Impacts . . . . . . . . . . . . . . . . . . 5-105.2.2 Exposure to Mercury Emissions and Health Impacts . . . . . . . . . . . . . . . . 5-12

5.3 Ecological Risks: Methodology and Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-14

5.3.1 Methodology for Assessing Ecological Risks . . . . . . . . . . . . . . . . . . . . . . 5-145.3.2 Ecological Risk Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-15

5.4 Other Socio-Economic Benefits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-15

5.4.1 Property Value Benefits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-165.4.2 Other Benefit Categories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-26

5.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-29

REGULATORY FLEXIBILITY ANALYSIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CHAPTER 6

6.1 Methodology for Identifying Small Entities by the Proposed Rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-2

6.1.1 Statutory Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-26.1.2 Methodology for Evaluating Combustion Units for

Regulatory Flexibility Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-2


6.2 Number of Combustion Facilities Defined As A SmallBusinesses and Likely Impacts of Compliance Withthe Proposed Rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-3

6.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-3

ENVIRONMENTAL JUSTICE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CHAPTER 7

7.1 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-17.2 Summary Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-17.3 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-2

APPENDIX A: CORE DATA INPUTS

APPENDIX B: BASELINE COST REPORT

APPENDIX C: WASTE MINIMIZATION REPORT

APPENDIX D: WASTE MANAGEMENT ALTERNATIVE REPORT

APPENDIX E: MODEL OUTPUT FOR MACT OPTION 1B

APPENDIX F: REGULATORY FLEXIBILITY DATA

APPENDIX G: MARKET IMPACTS ASSUMING ZERO PERCENT PRICEPASS-THROUGH SCENARIO

LIST OF EXHIBITS

Exhibit ES-1: Regulatory Options For Existing Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ES-2

Exhibit ES-2: Integration of Model Plants Dataand Combustion Unit-Specific Datain Economic Screens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ES-7

Exhibit ES-3: Sectoral Comparison of Impacts of Proposed MACT Option IB . . . . . . . . . . . . . . ES-9

Exhibit 1-1: Regulatory Options For Existing Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-3

Exhibit 1-2: Regulatory Options For New Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-5

Exhibit 2-1: Number and Sector of Combustion Facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-6

Exhibit 2-2: Quantities of Hazardous Wastes Combusted in 1991 . . . . . . . . . . . . . . . . . . . . . . . . 2-8

Exhibit 2-3: Industrial Sectors Generating Combusted Wastes . . . . . . . . . . . . . . . . . . . . . . . . . 2-10

Exhibit 2-4: Baseline National Emissions ofHAPs from Combustion Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-11

Exhibit 2-5: Baseline APCDs by Combustion Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-13

Exhibit 3-1: Overview of Compliance Cost Model Plant Analysis . . . . . . . . . . . . . . . . . . . . . . . 3-3

Exhibit 3-2: Emission Control DevicesAssigned in Model Plants Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-4

Exhibit 3-3: Estimated Per-Unit Annual Cost of ContinuousEmissions Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-6

Exhibit 3-4: Calculation of Total Compliance Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-11

Exhibit 3-5: Scaling Factors for Total Cost Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-12

Exhibit 3-6: Total Annual Compliance Costs (Assuming No Market Exit) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-14

Exhibit 3-7: Percentage of Units Requiring Control Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-16

Exhibit 3-8: Percentage of Compliance Costs by Control Device . . . . . . . . . . . . . . . . . . . . . . . 3-17

Exhibit 3-9: Average Total Annual Compliance Costs Per Unit . . . . . . . . . . . . . . . . . . . . . . . . 3-19

Exhibit 3-10: Effect of 25 Percent Feed Reduction on Total Annual National Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-20

Exhibit 3-11: Per-Unit Incremental Annual Costs for New Sources . . . . . . . . . . . . . . . . . . . . . . 3-22

LIST OF EXHIBITS(Continued)

Exhibit 4-1: Integration of Model Plants Dataand Combustion Unit-Specific Datain Economic Screens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-4

Exhibit 4-2: Baseline Cost Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-10

Exhibit 4-3: Baseline Operating Profits Per Tonof Hazardous Waste Burned . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-11

Exhibit 4-4: Combustion Cessation Analysis Using Break-Even Quantity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-20

Exhibit 4-5: Simplified Example of Determination ofNew Market Price for Combustion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-21

Exhibit 4-6: Unit Consolidation Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-23

Exhibit 4-7: Average Total Annual Baseline and Compliance Costs Per Ton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-29

Exhibit 4-8: New Compliance Costs as a Percentage of Hazardous Waste Burning Revenues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-30

Exhibit 4-9: Compliance Costs as a Percent of Baseline Costs . . . . . . . . . . . . . . . . . . . . . . . . . 4-32

Exhibit 4-10: Summary of Waste Minimization Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-34

Exhibit 4-11: Applicability of Alternative Combustion Technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-37

Exhibit 4-12: Quantity of Key Wastestreams that Could Potentially be Managed by each of the Currently Viable Alternative Technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-39

Exhibit 4-13: Cost Data on Most-Applicable Alternative Technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-40

Exhibit 4-14: Weighted Average Price Per Ton and Increase in Prices due to Assumed Price Pass-Through (Low Price Pass-Through) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-44

Exhibit 4-15: Percentage of Combustion Units Meeting Short Term BEQ After Consolidation (Low Price Pass-Through) . . . . . . . . . . . . . . . . . . . . . . 4-45

Exhibit 4-16: Percentage of Combustion Units Meeting Long Term BEQ After Consolidation (Low Price Pass-Through) . . . . . . . . . . . . . . . . . . . . . . 4-47


Exhibit 4-17: Number of Facilities Likely to Stop Burning Waste in the Short Term (Low Price Pass-Through) . . . . . . . . . . . . . . . . . . . . . . . 4-49

Exhibit 4-18: Number of Facilities Likely to Stop Burning Waste in the Long Term (Low Price Pass-Through) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-50

Exhibit 4-19: Change in Average Operating Profits Per Ton From the Proposed MACT in the Short Term(Low Price Pass-Through) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-51

Exhibit 4-20: Quantity of Waste that Could be Diverted from Facilities in the Short Term (Low Price Pass-Through) . . . . . . . . . . . . . . . . . . . . . 4-53

Exhibit 4-21: Weighted Average Price Per Ton and Increase in Prices Due to Assumed Price Pass-Through (High Price Pass-Through) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-54

Exhibit 4-22: Percentage of Combustion Units Meeting Short Term BEQ After Consolidation (High Price Pass-Through) . . . . . . . . . . . . . . . . . . . . . . 4-56

Exhibit 4-23: Percentage of Combustion Units Meeting Long Term BEQ After Consolidation (High Price Pass-Through) . . . . . . . . . . . . . . . . . . . . . . 4-57

Exhibit 4-24: Number of Facilities Likely to Stop Burning Waste in the Short Term (High Price Pass-Through) . . . . . . . . . . . . . . . . . . . . . . . 4-58

Exhibit 4-25: Number of Facilities Likely to Stop Burning Waste in the Long Term (High Price Pass-Through) . . . . . . . . . . . . . . . . . . . . . . . 4-59

Exhibit 4-26: Change in Operating Profits Per Ton in the ShortTerm from the Proposed MACT (High Price Pass-Through) . . . . . . . . . . . . . . . . . 4-61

Exhibit 4-27: Quantity of Waste That Could be Diverted From Facilities in the Short Term (High Price Pass-Through) . . . . . . . . . . . . . . . . . . . . 4-63

Exhibit 4-28: Total Annual Pre-Tax Compliance Costs After Unit Consolidations (Low Price Pass-Through) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-67

Exhibit 4-29: Total Annual Pre-Tax Compliance Costs After Unit Consolidations (High Price Pass-Through) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-68

Exhibit 4-30: Annual Pre-Tax Compliance Costs Per Ton After Consolidation (Low Price Pass-Through) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-69

Exhibit 4-31: Annual Pre-Tax Compliance Costs Per Ton After Consolidation (High Price Pass-Through) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-70


Exhibit 4-32: Number of Units Changing Model Plant Assignments After Waste Feed Reductions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-71

Exhibit 4-33: Revenue Losses Under Feed Reduction Compliance Option . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-72

Exhibit 4-34: Cost Impact of Removing Small Quantity Burner Exemption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-74

Exhibit 4-35: Sectoral Comparison of Impacts of Proposed MACT Option 1B . . . . . . . . . . . . . . 4-78

Exhibit 5-1: Total Compliance Costs by Type of HAP Controlled . . . . . . . . . . . . . . . . . . . . . . . 5-5

Exhibit 5-2: Reduction in Hap Emissions Under Each Option . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-6

Exhibit 5-3: Cost Per Ton of Hap Reduced . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-8

Exhibit 5-4: Comparison of Cost-Effectiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-9

Exhibit 5-5 Summary of Baseline Ecological Risk Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-16

Exhibit 5-6 Summary of Property Value Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-18

Exhibit 5-7 Property Value Impacts for Sample Counties/Facilities . . . . . . . . . . . . . . . . . . . . . 5-23

Exhibit 5-8 Property Value Impacts for Sample Counties/Facilities . . . . . . . . . . . . . . . . . . . . . 5-25

Exhibit 5-9 Total Annual Property Value Impacts (1 mile boundary) . . . . . . . . . . . . . . . . . . . . 5-25

Exhibit 6-1: Summary of Combustion Units Subject to Regulatory Flexibility Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-4

Exhibit 6-2: Detail on "Small" Combustion Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-5

Appendix F: Comparison of Facility and Parent Revenues and Employment to Small Business Thresholds

Appendix G: Market Impacts Assuming Zero Percent Price Pass-Through Scenario

DRAFT: November 13, 1995ES-1

EXECUTIVE SUMMARY

ES1.1 OVERVIEW

In May of 1993, the Environmental Protection Agency (EPA) introduced a draft WasteMinimization and Combustion Strategy to reduce reliance on the combustion of hazardous waste andencourage reduced generation of these wastes. Among the key objectives of the strategy is the reductionof health and ecological risks posed by the combustion of hazardous waste. As part of this strategy, EPAis developing more stringent MACT emissions standards for waste combustion facilities. These newMACT standards address a variety of hazardous air pollutants (HAPS), including dioxins/furans, particulatematter, mercury, semi-volatile and low-volatility metals, and chlorine. In addition, emissions of carbonmonoxide and hydrocarbons will be regulated as proxies for products of incomplete combustion (PICs).The rule sets emission levels for commercial incinerators, waste-burning cement kilns and lightweightaggregate kilns, and on-site incinerators. The proposed rule is scheduled for release in late 1995.

This Regulatory Impact Analysis (RIA) has been completed in accordance with Executive Order12866, which requires EPA to develop and submit to the Office of Management and Budget (OMB) an RIAfor any significant regulatory action. This document also fulfills the requirements of the RegulatoryFlexibility Act mandating that EPA evaluate the effects of regulations on small entities.

The RIA assesses the costs of the rule and the impacts that these costs would have on waste burningbehavior, and compares these costs to the benefits of the regulation. By evaluating a variety of regulatoryalternatives, EPA has identified pollutant control levels that provide cost-effective protection of publichealth and the environment. At present, the RIA addresses seven regulatory alternatives for existingcombustion sources (see Exhibit ES-1), although EPA is in the process of developing five additionaloptions that will be summarized in an addendum. EPA also evaluated three regulatory alternatives for newsources. In addition to insights on the costs and benefits of the proposed standards, the RIA also providesthe Agency with important information on how the proposed rule might affect the competitive dynamicsin combustion markets.

ES2.1 SUMMARY OF FINDINGS

This RIA provides estimates of the costs and benefits of EPA's proposed MACT standards forhazardous waste combustion facilities. The RIA analysis suggests that the rule may have substantialbenefits, including reductions in adverse health effects, reduced levels of toxic substances in ecosystems,and improvements in property values at homes located near combustion facilities. EPA's analysis of thecosts and economic impacts of the new standards indicates that, although a sizeable number of combustionfacilities may stop burning hazardous waste, the vast majority of them will remain viable productionfacilities. Most cement kilns and LWAKs that stop burning will be able to continue clinker and aggregateproduction by shifting to conventional or non-hazardous waste fuels. Decisions to halt waste burning areexpected to result in plant closures only at a few commercial incinerators burning small quantities of waste.Although many on-site incinerators are expected to close, industrial production should continue at thesesites because waste burning represents only a small fraction of total production costs. In addition, someon-site incinerators may continue to burn non-hazardous wastes. The remainder of this section summarizesthe central conclusions of the RIA.


Exhibit ES-1

REGULATORY OPTIONS FOR EXISTING SOURCES

Option Category (ng/dscm TEQ) (gr/dscf) (ug/dscm) (ug/dscm) (ug/dscm) (ppmv) (ppmv) (ppmv) (ppmv)Source D/F PM Hg SVM LVM HCl Cl CO HC2

Floor Levels Incinerators 0.5 0.015 30 60 80 25 1 100 20

Cement Kilns 0.5 0.03 40 60 80 60 1 NA 20

LWA Kilns 0.5 0.015 30 60 80 1300 2.5 100 20

MACT Option 1b: Basic Incinerators ATF 0.2 0.015 30 60 80 25 1 100 20Option (includes common,protective D/F stnd andcommon Hg standard)

Cement Kilns ATF 0.2 0.03 ATF 30 60 80 60 1 NA 20

LWA Kilns ATF 0.2 0.015 30 60 80 1300 2.5 100 20

MACT Option 1c: Basic Incinerators ATF 0.2 0.015 ATF 5 60 80 25 1 100 20Option (includes common,protective standards forboth D/F and Hg)


LWA Kilns ATF 0.2 0.015 ATF 5 60 80 1300 2.5 100 20

MACT Option 2a: Basic Incinerators ATF 0.2 0.015 ATF 5 60 80 25 1 ATF 50 ATF 5Option 1c + ImprovedCombustion Cement Kilns ATF 0.2 0.03 ATF 5 60 80 60 1 ATF 50 ATF 5

LWA Kilns ATF 0.2 0.015 ATF 5 60 80 1300 2.5 ATF 50 ATF 5

MACT Option 2b: Basic Incinerators ATF 0.2 0.015 ATF 5 60 80 25 1 ATF 50 ATF 5Option 1c + ImprovedCombustion (W/CKexemption)



MACT Option 3: Basic Incinerators ATF 0.2 0.015 ATF 5 60 80 25 1 100 20Option 1c + CommonStandards Cement Kilns ATF 0.2 ATF 0.015 ATF 5 60 80 60 1 NA 20

LWA Kilns ATF 0.2 0.015 ATF 5 60 80 ATF 60 ATF 1 100 20

MACT Option 4: Basic Incinerators ATF 0.2 0.015 ATF 5 60 80 25 1 ATF 50 ATF 5Option 1c + CommonStandards + ImprovedCombustion

Cement Kilns ATF 0.2 ATF 0.015 ATF 5 60 80 60 1 ATF 50 ATF 5

LWA Kilns ATF 0.2 0.015 ATF 5 60 80 ATF 60 ATF 1 ATF 50 ATF 5


! Total Costs of the Rule Assuming No Market Exit range from $189 million to$561 million per year, depending on the MACT alternative. This is an upperbound cost estimate, as a number of facilities will choose to stop burninghazardous wastes rather than incur the rule's compliance costs.

! Total Costs of the Rule Allowing for Market Exit range from $140 to $176million per year for most MACT options when EPA assumes firms will have littleability to pass through compliance costs to generators ("low price pass-through").Options requiring PIC control generate much higher compliance costs, rangingfrom $191 to $285 million per year. Since capital and operating expenditures aretax deductible, the tax-adjusted cost of the rule for individual generators will besomewhat lower, although the exact cost will depend on the marginal tax rates ofthe individual generators.

! Most Compliance Costs Are Incurred to Reach the Floor. With the exceptionof MACT options requiring PIC control (2a, 2b, and 4), most of the costs of therule and decisions to halt waste burning occur in reaching the MACT floor.

! Market Exit in the Short-Term. EPA estimates that, if facilities have littleability to pass on compliance costs to generators in the form of higher combustionprices, 58 to 71 facilities (depending on MACT option) will cease hazardous wasteburning. Nearly 90 percent of the exit is in the on-site incinerator sector forregulatory alternatives that do not include PIC controls.

! Market Exit in the Long-Term. As waste burning capital must be replaced, anadditional 27 to 34 facilities (depending on MACT option) are expected to stopburning hazardous wastes. Again, these are mostly on-site incinerators.

! Capacity Utilization Drives Exit. Facilities that cease burning wastes aregenerally those that burn small quantities of waste now, and therefore face highcompliance costs per ton of waste burned. For this reason, EPA estimates thatdecisions to stop burning waste will divert only between three and 10 percent ofwaste quantities currently burned. Combustion units that adequately utilizeexisting waste burning capacity will remain viable.

! High Price Pass-Through Scenario. If combustion facilities are able to passthrough a greater portion of compliance costs to generators, a lower the numberof facilities are expected to exit the waste burning market. While this reduceswaste diversions from facilities exiting the market to only two to three percent,greater pass-through does increase the social costs of the rule somewhat.

! Operating Profits. Most combustion facilities will experience reduced operatingprofits from hazardous waste burning as a result of the rule. Under EPA's lowprice pass-through scenario, percentage declines in profit margins are largest in thecement and LWAK sectors. Nonetheless, EPA's analysis suggests that, on


average, cement kilns and LWAKs will remain profitable, with post-complianceoperating margins on waste burning that are still higher than those for commercialincinerators.

! Waste Minimization and Non-Combustion Alternatives. The proposed rulewill provide greater incentives for the adoption of waste minimization and otheralternatives to combustion, especially in the on-site incinerator sector. For manyon-site incinerators, compliance with the proposed rule is predicted to increasewaste burning costs substantially; EPA expects that many generators will findwaste minimization a cheaper compliance option when compared with thecommercial combustion alternative. The Agency, however, is unable to predictexactly how much waste minimization will occur because of the complex natureof existing institutional barriers to further waste reduction. To further encouragewaste minimization, EPA is requesting comments on a number of additionalprovisions in the proposed rule. First, the Agency may require that on-sitecombustors better characterize their waste streams as a way to improve theapplicability of waste minimization plans. Second, in return for specific wasteminimization commitments, the Agency may allow combustors additional time tocomply with the new regulations or delay calling-in existing permits. While suchexceptions will be determined on a case-by-case basis, they could increase theeconomic viability of particular waste minimization options.

! Compliance Costs Per Ton Similar for New and Existing Sources. Thissuggests that the proposed rule will not create any substantial barriers to entry inthe hazardous waste combustion industry.

! Cost Effectiveness. Average compliance costs per ton of HAP controlled rangefrom about $5,000 to $8,500. Costs are lowest under the floor and options 2A and4. Control of dioxin, mercury, and semi-volatile metals drive compliance costsunder most MACT options. Control expenditures vary little across MACT optionsfor dioxin/furans; variability is greater for mercury controls.

! Benefits: Population risks are considerable for dioxin and mercury in the pre-regulatory baseline. EPA estimates that hazardous waste burning sourcesrepresent about nine percent of total anthropogenic dioxin emissions and aboutfour percent of total anthropogenic mercury emissions in the U.S. Under theproposed MACT standard, reductions in dioxin and mercury emissions fromhazardous waste burning sources are significant. EPA expects that thesereductions, in conjunction with reductions in emissions from other dioxin andmercury-emitting sources, will help reduce dioxin and mercury levels over timein foods used for human consumption and therefore, reduce the likelihood ofadverse health effects, including cancer and neurological effects in adults, anddevelopmental abnormalities in children. Screening analyses also suggest thatemissions reductions may lead to property value increases, an alternative measureof benefits. While a precise estimate is not possible due to numerousuncertainties, property value increases have the potential to be substantial underthe rule.


! Effects of Other EPA Initiatives. Other pending EPA initiatives couldsubstantially alter the results presented in this RIA. Most importantly, the MACTrule governing emissions from non-hazardous waste burning cement kilns couldreduce the incremental cost of burning waste in kilns, providing incentives formore kilns to enter the waste burning market. On the other hand, newmanagement standards for cement kiln dust (CKD) could reduce the overallcompetitiveness of the kilns if waste burners must manage greater quantities ofCKD than non-burners. Furthermore, the exemption of on-site boilers from thisproposed rule could lead to increased use of boilers, reducing waste quantitiesflowing to the sectors analyzed here. Finally, a possible exemption for "clean"hazardous waste fuels (i.e., those comparable to conventional fuels in theiremissions profiles) from RCRA would probably shift the burden of newcompliance costs onto the more highly contaminated solids, sludges, and liquids.The details of this provision continue to be worked out; therefore, EPA is unableto provide a quantitative assessment of how they would alter the estimatespresented in the RIA for this rule.

Overall, the RIA analysis suggests that the benefits of the rule are potentially substantial. While thesecannot be fully monetized in units comparable to the costs, the analysis suggests that MACT standards forhazardous waste combustion facilities are necessary to provide significant reductions in emissions andexposures from this sector. A screening analysis of property value impacts of the new standards alsosuggests that benefits could be substantial. Given the severe nature of the adverse health effects causedby dioxin and mercury (cancer, developmental effects in children, and severe neurological effects inadults), EPA believes that substantial reductions in emissions of these pollutants from hazardous wasteburners under the MACT standard justifies moving ahead with the proposed above the floor (ATF) option.

ES3.1 ECONOMIC IMPACT ASSESSMENT

EPA's proposed MACT standards will increase the cost of hazardous waste combustion across theindustry, and will lead some facilities to stop burning hazardous wastes altogether. The market impacts,however, are expected to be small. Only a few facilities are expected to stop burning wastes in thecommercial sectors. While more on-site incinerators will cease waste burning, they currently tend to burnsmall quantities of hazardous waste. Waste diversions from combustion are likely to be minor as a result.The rule is expected to cost between $140 and $187 million per year, although costs jump to a high ofbetween $285 and $405 million per year if PIC controls at kilns are required. EPA does not expect the ruleto trigger large changes in the competitive balance between these sectors.

The quantities of hazardous waste currently combusted are the primary determinant of whethercombustion facilities will stop burning. Much of the market exit in all four sectors is driven by the smallquantities of waste over which facilities are able to spread the fixed costs of compliance with the newstandards. Since the facilities that stop burning hazardous waste tend to be those currently burning smallquantities, EPA estimates that only about three percent of the tonnage currently combusted will be divertedto other combustion sites or to non-combustion management alternatives.

Most of the costs of the rule, and market exit from the rule, occur in reaching the proposed MACTfloor. With the exception of options that require PIC controls (2a, 2b, and Option 4), the above-the-flooroptions do not noticeably change the impacts of the rule. Under proposed MACT options requiring PIC


controls, compliance costs per ton rise dramatically for cement kilns and LWAKs, suggesting that moresubstantial changes in the combustion marketplace would occur.

Uncertainty about the cost and availability of waste minimization and non-combustion wastemanagement alternatives for generators makes it difficult to predict how high combustion prices could risebefore generators switched substantial waste quantities to non-combustion alternatives. Although EPAbounds this uncertainty by including a zero-percent price pass-through scenario, the impact of price changeson market exit is not large in most sectors. The pricing assumptions have a much more direct effect on theprofitability of the units that remain open after complying with the rule.

Since firms that close combustion units in the face of the new rule do not incur complianceexpenditures for those units, the estimate of the total costs of the rule excludes the units likely to stopburning hazardous wastes. The total cost estimate includes compliance costs for all combustion units thatcontinue to burn hazardous wastes, plus any cost increases from having to manage wastes from closedfacilities in a more expensive manner. For all non-PIC proposed MACT options, total costs of the rule are22 to 29 percent lower than total engineering costs that assume all units comply with the rule. The costsfor PIC options decline by almost 50 percent once market exit is incorporated.

ES3.1.1 Economic Impact Methodology

To estimate compliance costs associated with the MACT proposal, EPA used a model plantsapproach augmented by certain plant-specific and industry-specific data. The pre-MACT baseline cost ofcombustion was estimated using one set of model plants, based on existing plant type, size, and installedair pollution control devices (APCDs). Separate model plants for compliance costs were also created, againbased on plant type and size, as well as on new APCDs required to comply with the proposed rule. Thesetwo sets of model plants form the basis for all the economic impact analysis in the RIA. Additional coststhat applied across all combustion units, such as for continuous emission monitors, were added to this total.

As illustrated in Exhibit ES-2, model plant data were combined with plant- and industry-specificdata on quantities and prices of waste burned by type. From these core data, EPA estimated compliancecosts, waste burning revenues, and energy savings (where applicable) for each combustion unit analyzed.Compliance costs were then evaluated against a variety of economic measures to assess the impact of therule on combustion markets. Basic screening measures included compliance costs per ton of waste burned,compliance costs as a percentage of waste burning revenues, and compliance costs as a percentage ofbaseline costs to burn hazardous waste. Each screen provides useful information on the relative magnitudeof compliance costs under various regulatory scenarios and helps the Agency gauge the differential impactof the proposed MACT standards across combustion sectors.


While the basic screens provide rough indicators of how the effects of the rule vary acrossindividual combustion units and combustion sectors, EPA used a breakeven quantity approach (BEQ) asa more precise indicator of when specific combustion units would no longer be profitable as hazardouswaste burners. The BEQ, a planning tool commonly used by businesses, measures the quantity of wastethat a combustion unit would have to burn to cover the costs of operation. EPA used estimates ofbreakeven quantity to assess the likelihood that combustion facilities will stop burning waste in the faceof increased compliance costs by comparing the BEQ to data on the actual tons burned. Where multiplecombustion units at a single site failed to meet BEQ, waste flows were consolidated into fewer combustionunits under the assumption that a rational waste combustor would undertake such an action to maximizeprofits.

The BEQ is the basis for the Agency's estimates of market exit in both the short- and long-term.In addition, because the BEQ changes when prices change, EPA was able to model the effect of threedifferent pricing scenarios on estimated market exit. Each scenario reflected different assumptions aboutthe market price at which waste minimization and/or other alternatives to combustion would becomecompetitive. These pricing scenarios enabled the Agency to bound the estimated impacts of the proposedrule despite uncertainty about the economics of alternatives to combustion.

While providing a general indication of each combustion facility's likely response to the newstandards, the BEQ measure does have limitations. The first important limitation involves the quality ofdata on tons burned, type of waste burned, and prices charged. BEQ results are especially sensitive to thequantity data. The second limitation is inherent in the design of the breakeven measure itself. Evencombustion units that meet BEQ may earn profits too low to justify continued waste burning. EPA hastherefore assessed changes in operating profits due to the rule. The Agency invites industry to providesupplemental data to address either limitation in order to enhance EPA's evaluation of impacts of the ruleon industry or sector profitability.

ES3.1.2 Differential Impacts of the Rule by Market Sector

Although the rule will not cause large changes in sectoral competition, not all market participantswill be affected in the same way. This section summarizes the key impacts on the four combustionsegments, as well as on fuel blenders and generators. A comparison of important measures of impactacross sectors is presented in Exhibit ES-3.

ES3.1.2.1 Cement Kilns

Cement kilns currently have a baseline cost advantage over commercial incinerators due to theirability to use capital equipment (such as the kiln) for joint cement production/hazardous waste destruction.While the proposed MACT standards will greatly change the baseline cost of burning waste at a cementkiln, the existing cost advantage is large enough that cement kilns are expected to remain the lowest-costcombustors under all proposed MACT scenarios.

Kilns likely to exit the market are those that tend to burn little waste now. Under proposed MACToption 1b, between two and three facilities currently burning hazardous wastes are expected to stop doingso. This remains at three facilities if prices could not rise at all. If kilns are required to install PIC controls(as under proposed MACT options 2a and 4), and are not successful in passing the full costs of thesecontrols to generators in the form of higher prices, ten kilns are expected to stop burning hazardous wastesover the long-term (17 if prices couldn't rise at all). This is a more likely scenario than full price pass-through. Although the kilns would have the lowest total combustion costs per ton even with PIC controls,

Kilns burn primarily high-Btu liquid solvent wastes. Solvent recycling would be a cost-effective1

alternative, limiting the degree to which kilns could increase prices. However, solid hazardous wastes aresuspended in the solvents prior to burning. Solvent recycling would eliminate this disposal outlet for solids,suggesting that prices on liquids with suspended solids could rise much more than prices on clean solvents.


the Agency believes that a tripling of combustion prices at the kiln probably would cause generators toadopt alternative management methods.1

Despite the low expected market exit, kilns remaining in the market would experience a substantialdecline (on the order of 25 to 30 percentage points) in the profitability of hazardous waste combustion.This decline will make it more difficult for kiln operators to cross-subsidize cement production withhazardous waste profits. While strong cement markets now protect many of these marginal kilns, reducedhazardous waste profits could accelerate plant closures during the next construction industry downturn.EPA currently does not have adequate information on marginal plants to assess the likelihood of suchclosures. Despite declines in waste burning operating profits due to the new standards, however, most BIFsare expected to maintain operating margins higher than those of commercial incinerators.

Exhibit ES-3

SECTORAL COMPARISON OF IMPACTS OF PROPOSED MACT OPTION 1B

Cement Commercial Aggregate On-SiteKilns Incinerators Kilns Incinerators

Lightweight

Total Compliance Costs ($Millions)- Zero pass-through $51 $19 $3 $72- Low pass-through $56 $19 $3 $74- High pass-through* $60 $19 $3 $77

Average Compliance Costs Per Tonof Hazardous Waste Burned- Zero pass-through $59 $30 $44 $64- Low pass-through $62 $30 $44 $63- High pass-through* $69 $30 $44 $90

Number of Combustion FacilitiesLikely to Stop Burning HazardousWastes in the Long-Term**- Zero pass-through 3 6 1 89- Low pass-through 3 6 1 82- High pass-through 2 6 1 72

Notes:* Costs rise in the high pass-through scenario because some units with higher compliance costs

remain viable.** Excludes facilities not burning waste in the baseline. Exit over long-term capital cycle;

immediate market exit would be lower. Due to the large number of very small on-site burners,on-site incinerators burning less than 50 tons of hazardous waste per year excluded fromcalculation. For all other sectors, only units burning zero tons per year were excluded.

This market erosion may not necessarily occur in a linear fashion. For example, some kilns may not2

refocus on hazardous waste combustion expansion until the demand for cement slows.


The market behavior of waste-burning cement kilns in the short-run and the long-run is quitesimilar. Because the kilns have little in the way of baseline capital equipment dedicated solely to wasteburning, the long-term BEQ is quite similar to the short-term BEQ.

ES3.1.2.2 Commercial Incinerators

Commercial incinerators face lower compliance costs per ton of waste burned than do cement kilns.As a result, commercial incinerators that remain in operation will become somewhat more competitive withthe kilns. The dollar value of this benefit, however, is about $40 per ton under all proposed MACT optionsnot requiring PIC controls for kilns, and is therefore not expected to affect market competition significantly.EPA estimates that commercial incinerators will continue to incur average costs per ton of waste burnedthat are substantially higher than BIFs. Though their ability to handle waste streams that cannot currentlybe burned in kilns provides the incinerators with a protected niche, the longer-term erosion the commercialincinerators' market is likely to continue.2

Facility closure in the commercial incinerator sector is driven almost entirely by waste quantitiesburned; low quality burners are much more likely to close. Four of the permitted facilities currentlyburning wastes are expected to close under both the high and the low price pass-through scenarios and areprobably marginal even in the current market. Long-term closures increase by two facilities in both pricepass-through scenarios for the alternatives without PIC controls. This increase is driven by the high capitalcosts that must be recovered over the capital replacement cycle in order for incinerators to stay in themarket.

Commercial incinerators also experience a decline in operating profit margins — of between 5 and20 percent in the low price pass-through scenario. As the declines are smaller than for cement kilns, therule will improve the competitive position of commercial incinerators relative to kilns. EPA, however,does not anticipate that the changes are large enough to cause substantial changes in competitive dynamics.

ES3.1.2.3 Lightweight Aggregate Kilns

LWAKs face average compliance costs per ton of waste burned (before consolidation) up to twiceas high as those for cement kilns and commercial incinerators. This differential is overstated, however,because over 30 percent of the permitted LWAKs do not currently burn wastes in their kilns. Once non-viable units are removed from the market, average compliance costs per ton drop below costs to cementkilns for most MACT options. Furthermore, EPA estimates that LWAKs have a baseline cost advantageover commercial incinerators similar to that enjoyed by kilns. Before and after the rule, LWAKs retaina strong competitive position.


Kilns with little or no waste now will most likely stop waste burning in the face of the proposedMACT standards. The substantial required investments in pollution control capital cannot be spread overthe small tonnage at these facilities, causing the units to stop burning hazardous wastes. Under the non-PICcontrol alternatives, two LWAKs are likely to exit the market, only one of which is actively burning wastes.As with cement kilns, exit increases under proposed MACT options requiring PIC controls unless thesecosts can be passed on to generators in the form of higher prices. Overall, LWAKs currently burning largerquantities of hazardous waste will remain strong competitors in the hazardous waste combustion marketfor both the short- and the long-run. While operating profit margins for remaining facilities do decline,margins after compliance with the rule are expected to remain higher than those in the commercialincinerator sector.

ES3.1.2.4 On-site Incinerators

On-site incinerators fall into two categories. At one end of the spectrum are facilities with highcapacity utilizations and burning large quantities of waste. These facilities are able to spread fixedinvestments so that costs per ton of waste combusted are competitive with commercial alternatives. At theother end of the spectrum are a large number of incinerators that burn extremely low tonnages; it is thesefacilities that lead to the very high compliance costs per estimated ton of waste burned for this sector. Tothe extent that these units also burn large quantities of non-hazardous wastes, some may remain viableunder the proposed MACT standards. More likely, however, is that the units will stop burning hazardouswaste even if non-hazardous waste continues to be burned. This is because the BEQ on compliance costsalone frequently exceeds the tons of hazardous waste currently burned at the units.

Market exit in this sector is expected to be substantial — between 44 and 61 facilities in the short-term (assuming low price pass-through). As the incinerator capital itself requires replacement, over one-half of the on-site incinerators actively burning are expected to stop burning hazardous wastes. Sincequantities burned at each unit are quite low, the overall impact of these shutdowns on the combustionmarket is expected to be minimal. Units that remain in the market, however, maintain operating profitmargins similar to, or better than, commercial incinerators under most MACT options.

ES3.1.2.5 On-site Boilers

On-site boilers are not covered under the proposed MACT standards. To the extent that the samewastes can be combusted in both on-site boilers and on-site incinerators, the rule makes boilers more cost-effective because they do not have to invest in new controls. Therefore, on-site boilers could provide alower-cost outlet for some of the wastes diverted from on-site incinerator closures (thereby reducing thecost of the rule). Alternatively, the existence of a boiler option could accelerate the closure of incineratorsas firms divert wastes even from incinerators that appear capable of meeting their BEQ. Data were notavailable that to allow EPA to determine the magnitude of these impacts.

EPA is considering exempting the combustion of clean hazardous waste fuels from regulations governing3

hazardous waste combustion in general on the grounds that there is little difference between them andconventional fuels such as coal and oil.

Blending operations at commercial incinerators are often done on-site by incinerator employees, whereas4

blending at cement kilns is often run by outside firms.


ES3.1.2.6 Hazardous Waste Generators

Hazardous waste generators are likely to see price increases for combusted waste streams. Withall combustion sectors facing increased costs, and with the compliance costs per ton that are not radicallydifferent across commercial sectors, generators will have little flexibility to seek out a lower costcombustion sector. The size of the likely price increases is difficult to determine, and will differ by thetype of waste. EPA anticipates that generators of clean solvents and lean waters will face lower priceincreases due to the availability of non-combustion alternatives. High-Btu liquids without suspended solidscan be reclaimed, or possibly burned in exempt facilities, both of which would constrain price increasesthat could be passed on to generators of the liquids. Land-ban solids and sludges could face more3

substantial increases.EPA does not have sufficient information on specific generating sectors to evaluate the impact of

price increases on generator processes more specifically. The Agency, however, does expect that a numberof waste minimization and non-combustion alternatives are available in the long-term should combustionprices rise significantly. For this reason, EPA anticipates that price increases of between $20 and $90 perton are more likely than larger increases. These changes would increase waste management costs forgenerators, although EPA has not analyzed the impact this could have on the prices of products producedby these generators.

ES3.1.2.7 Fuel Blenders

Fuel blenders serve as intermediaries between generators and combustors in both the commercialBIF and commercial incinerator sectors. To the extent that the proposed MACT changes the combustion4

demand for any one of these parties, blenders would need to react.

The RIA indicates that the impacts of the rule on blenders are likely to be relatively small. Veryfew combustion facilities can avoid capital equipment purchases for MACT controls by reducing the metalsor chlorine in the hazardous waste fuels they burn, even if the new formulations are assumed to cost thesame as the old. As a result, blenders are unlikely to be called on to change their fuel blends by more thana handful of customers. Waste diversions from closed facilities are also expected to be quite small. Whilesome localized blender operations may need to ship wastes to different outlets, the magnitude of thesechanges nationally is unlikely to be large.

One exception involves proposed MACT options requiring PIC controls at commercial BIFs. Ifthe cost of compliance makes a higher number of kilns uncompetitive (a distinct possibility), blenders willlose their key waste outlet. Alternatively, PIC controls could bifurcate the BIF market, with liquidsexempted by clean fuels regulations burned in non-waste burning BIFs (or recycled), while solids areburned in commercial incinerators. This could drastically reduce the demand for fuel blending services.


ES3.1.3 Uncertainties With Economic Impact Assessment

In addition to the data limitations mentioned in the BEQ discussion, two additional caveats areworth noting:

ES3.1.3.1 Cross Subsidies to Cement Manufacture

The RIA treats combustion activities as an independent profit center for BIFs. That is, combustionunits meeting BEQ on waste burning are assumed to remain in operation. EPA has not evaluated the casewhere profits from waste burning are used to cross-subsidize inefficient cement production activities.While such behavior could not be sustained over a long period of time (such units would be out-competedby units that efficiently produce cement and burn hazardous wastes), it can exist in the short-term, and maybe a real concern for a small number of older, wet process kilns in the cement industry.

The Agency does not expect this omission to greatly alter the short-run conclusions of this analysisin any substantial way because cement markets are extremely healthy now and operating profits on wasteburning remain quite strong even after the rule. Nonetheless, EPA invites industry to submit supplementaldata that would enable the Agency to analyze whether reduced operating profits in a market with cross-subsidies for cement manufacture would lead to additional plant closures.

ES3.1.3.2 Joint Use of On-Site Incinerators

Data available to EPA suggests that many on-site incinerators burn very small quantities ofhazardous waste. This yields large baseline capital costs per ton of waste burned. To the extent thatindustry also burns large quantities of non-hazardous waste in these incinerators, EPA's analysis wouldoverstate the baseline costs of burning, thereby overstating expected market exit somewhat. The extremelyhigh new compliance costs per ton of waste burned (properly attributed entirely to the hazardous wastefraction) are usually sufficient to trigger market exit in the on-site sector, even if baseline costs are assumedto be zero. For this reason, the Agency does not anticipate that better data on the joint use of on-siteincinerators for non-hazardous waste would change the conclusions of this RIA. Nonetheless, EPA invitesindustry to provide additional data on this issue if it is in disagreement.

ES4.1 BENEFITS OF THE RULE

ES4.1.1 Cost Effectiveness

EPA has developed a cost-effectiveness measure that examines cost per unit reduction of emissionsfor each HAP. The two analytic components of this measure are: (1) estimates of expenditures per HAPfor each regulatory option; and (2) estimates of emissions reductions under each regulatory option. EPAhas developed a simplified method that distributes expenditures to each HAP based on the cost of thetechnology that controls the HAP. In cases where the technology controls more than one HAP, costs aresplit based on the relative percent reduction required. Total mass emission reductions are estimated bysumming emissions estimates for all facilities. The emission reductions are then calculated as thedifference between the baseline total emissions and the emissions under the regulatory option.


As would be expected, control of dioxin and mercury accounts for the greatest share ofexpenditures, and the average costs (expenditures per unit reduced) increase with more stringent controls.Across all HAPs, cost per ton of emissions reduced ($5,000 to $8,500 per ton) is somewhat above, butgenerally comparable to, other combustion rules for municipal and medical waste combustors.

ES4.1.2 Human Health Risk

Population risks are considerable for dioxin and mercury in the pre-regulatory baseline. EPAestimates that hazardous waste burning sources represent about nine percent of total anthropogenic dioxinemissions and about four percent of total anthropogenic mercury emissions in the U.S. Human exposureto dioxin may cause as many as 600 cancer cases each year in the United States. For mercury, health risksinclude severe neurological effects in the offspring of exposed pregnant women and neurological effectsin adults. Adverse health effects are most likely for persons who eat large amounts of freshwater fish, suchas recreational or subsistence anglers, because exposures to mercury through this pathway can besubstantial.

Under the proposed MACT standard, reductions in emissions from hazardous waste burningsources are significant. At the floor and proposed above-the-floor levels, dioxin emissions from hazardouswaste combustion will be reduced by approximately 78 percent and 89 percent, respectively. Thesereductions correspond to an approximate three percent reduction in total estimated anthropogenic dioxinemissions in the U.S. Under the proposed MACT standard, mercury reductions also are substantial. EPAestimates that mercury emissions from hazardous waste burning sources will be reduced to 3.8 Mg per yearat the proposed floor levels and to 2.0 Mg per year at the proposed above the floor standard. Thesereductions correspond to approximately a three percent reduction of total anthropogenic mercury emissionsin the U.S. EPA expects that the decrease in dioxin and mercury levels from hazardous waste combustionsources, in conjunction with reductions in emissions from other dioxin and mercury-emitting sources, willhelp reduce dioxin and mercury levels over time in foods used for human consumption and therefore,reduce the likelihood of cancer and neurological effects in adults, and developmental abnormalities inchildren.

ES4.1.3 Ecological Risk

Emissions from waste burning may also affect terrestrial and aquatic ecosystems in the areasaround combustion facilities. As an initial screen for assessing risk to aquatic ecosystems, EPA comparedthe concentration of pollutants in surface water to aquatic toxicity criteria.

The screening analysis of ecological risks considers modeled watershed concentrations relative torisk-based ambient water quality criteria. The analysis suggests that water quality criteria may be exceededin the most vulnerable watersheds around waste combustion facilities, particularly cement kilns. Thisgenerally occurs, however, only under high end emission assumptions. These risks are expected to beeliminated under all the proposed regulatory alternatives.


ES4.1.4 Property Value Benefits

EPA examined potential property value impacts introduced by hazardous waste combustionfacilities. A literature review of property value studies provided information on the impact thatincinerators, landfills, and waste sites have on the value of surrounding properties. These impacts aregenerally reflected as an increase in housing prices per unit increase in distance from the waste site. EPAcombined this premium information with data on housing densities around combustion facilities to developa screening estimate of annual property value damages that could be eliminated by emissions reductionsat all combustion facilities.

EPA developed a property value screening analysis that uses distance from a combustion facilityas a proxy for emissions reduction. The basis for the analysis is a recent study of property value effectsaround a municipal waste incinerator; this study demonstrated how housing prices increase with distancefrom the facility. Sensitivity analysis of two key parameters -- the boundary distance within which thepremium applies and the mileage distance used as a proxy for emissions reduction -- yields a wide rangeof benefit estimates. If we assume that only those homes within one mile of the facility are affected andif the reduction in concentration is equivalent to a 0.25 mile shift away from the facility, the resultingbenefit is approximately $12 million per year. Alternatively, if all homes within 3.5 miles are affected andthe emissions reduction is equivalent to moving a house 2.5 miles further from the facility, we obtain anannual benefit of $840 million. To know whether such benefits occur, however, requires more knowledgeabout how housing prices respond to changes in air pollution. While the screening analysis performed hereallows EPA to conclude that property value damages are potentially significant, this central question canbe adequately addressed only through further research.

ES5.1 REGULATORY FLEXIBILITY AND ENVIRONMENTAL JUSTICE

The proposed rule is unlikely to adversely affect small businesses for two important reasons. First,few combustion units are owned by businesses that meet the Small Business Administration (SBA)definition as a small business. Specifically, available data allow the Agency to identify only eight smallentities out of the more than two hundred EPA-listed combustion units burning hazardous waste.Furthermore, while over one-half of those that are considered small have a relatively few employees,annual sales for these facilities are in excess of $50 million per year.

Second, facilities most hurt by the rule tend to be those that burn very little waste and hence facevery high costs per ton burned. The screening analyses and breakeven quantity analyses used to evaluatethe economic impact of the proposed rule demonstrate that tons burned rather than firm size are the primarydeterminant of market exit. Although EPA does not have data on the waste combusted at every small unit,those that have high capacity utilizations will face relatively small cost increases per ton. Those that burnvery little waste in their existing units will close rather than comply with the proposed rule. Since their lowquantities led to the closure, it follows that their cost of off-site treatment will also be low.

To evaluate the environmental justice implications of the proposed rule, EPA looked at whetherminority percentages within a one mile radius of the waste burning facility were significantly different fromthe minority percentages in the county as a whole. EPA analyzed all hazardous waste burning cementplants as well as a sample of incinerators. In neither case did EPA find that plants were locateddisproportionately near populations with higher minority densities than the surrounding area.


ES6.1 SUMMARY

EPA's analysis of the proposed rule indicates that some combustion facilities may experience asubstantial change in the cost of burning waste, but that this change is likely to have a limited impact oncombustion markets. In terms of effects on waste-burning cost structure, cement kilns and lightweightaggregate kilns (LWAKs) are most affected by the regulation. This is primarily a product of their relativelylow baseline costs of burning, meaning that incremental compliance costs represent a large increase in theiroverall cost of burning waste. For incinerators, compliance costs are lower, represent smaller additionsto baseline costs, and change little across regulatory options. The analysis concludes that cement kilns havethe lowest waste burning costs even after regulation, and so will continue to have the greatest leverage toincrease prices.

To the extent that compliance costs cannot be passed through to generators and fuel blenders, theprofitability of waste burning in kilns will fall. Nonetheless, waste burning kilns are expected to havehealthy operating profit margins after the rule. Market exit in all sectors is concentrated among facilitiesthat burn small quantities of hazardous waste. While as many as 98 combustion facilities may stop burninghazardous wastes as a result of the proposed MACT options, the small quantities these facilities burnsuggest that market dislocations will be minor.

Overall, the social costs of the rule are balanced by a set of potentially substantial benefits. Giventhe severity of the potential adverse health effects from dioxin and mercury (cancer, developmental effectsin children, and severe neurological effects in adults), EPA believes the substantial reductions of thesepollutants from hazardous waste burning sources under the MACT standard justifies moving ahead withthe proposed ATF option. An alternative way of valuing benefits is the potential increase in propertyvalues around closed or more stringently regulated combustion facilities. The fact that this approach alsosuggests potentially substantial benefits strengthens EPA's belief that the costs of moving forward with theproposed ATF option are justified.

DRAFT: November 13, 19951-1

INTRODUCTION AND REGULATORY OPTIONS CHAPTER 1____________________________________________________________________________________

1.1 INTRODUCTION

In May of 1993, the Environmental Protection Agency (EPA) introduced a draft WasteMinimization and Combustion Strategy to reduce reliance on the combustion of hazardous waste andencourage reduced generation of these wastes. Among the key objectives of the strategy is the reductionof the health and ecological risks posed by the combustion of hazardous waste. As part of this strategy,EPA is developing more stringent emissions standards for combustion facilities. EPA is proposing revisedemissions standards for a number of hazardous air pollutants (HAPs) from three categories of hazardouswaste combustion facilities:

! Cement kilns;

! Lightweight aggregate kilns;

! Incinerators, both commercial and on-site.

In coordination with other EPA offices, the Office of Solid Waste has developed this analysis toevaluate the economic impact of the proposed regulatory options. The EPA's proposed rule is currentlyscheduled for release in late 1995. This analysis assesses the costs of the rule and the impacts that thesecosts would have on waste burning behavior, and compares these costs to the benefits of the regulation.This economic impact analysis has been completed in accordance with Executive Order 12866, requiringEPA to develop and submit to the Office of Management and Budge (OMB) a regulatory impact analysisfor any significant regulatory action. This document also fulfills the requirements of the RegulatoryFlexibility Act (at 5 U.S.C. 601 et seq.) mandating that EPA evaluate the effects of regulations on smallentities.

The remainder of this chapter provides a discussion of the various regulatory options that EPA isconsidering under the combustion technical standards rule and a review the organization of this document.

The MACT pool must include no fewer than five units.1

CO and HC are regulated as surrogates for products of incomplete combustion (PICs).2


1.2 REGULATORY OPTIONS

The proposed technical standards rule adopts a "maximum achievable control technology" (MACT)approach for establishing emissions standards. Under this approach, emissions concentration limits aredeveloped as a function of control technologies currently in place. Specifically, a MACT standard settingexercise rank orders existing sources by emissions and requires control levels equal to or better than thebest performing regulated sources. EPA may require more stringent controls -- referred to as "above-the-1

floor" controls -- if needed to ensure protection of human health and the environment or in the interest ofestablishing common standards across combustor types. In this rule, floor and "above-the-floor" levelswere developed for nine hazardous air pollutants (HAPs):

! dioxin/furan;

! particulate matter;

! mercury;

! semi-volatile metals (SVM);

! low volatile metals (LVM);

! hydrochloric acid (HCl);

! chlorine (Cl );2

! carbon monoxide (CO); and

! hydrocarbons (HC).2

For more detail on the specific methodology used in developing floor and "above-the-floor" control levels,the reader should refer to the preamble to this rule.

EPA is considering seven different regulatory MACT options for existing sources. Exhibit 1-1summarizes these options. As shown, the regulatory options begin with standards based on the MACTfloor for each HAP and then generally increase in stringency. Several HAPs -- dioxin, PM, and mercury --vary the most between options. Other HAPs -- HCl, Cl , CO, and HC -- vary only for the more stringent2

options. Semi-volatile and low volatile metals do not change under any of the regulatory options. Theregulatory options can be described as follows:


Exhibit 1-1

REGULATORY OPTIONS FOR EXISTING SOURCES


Floor Levels Incinerators 0.5 0.015 30 60 80 25 1 100 20

Cement Kilns 0.5 0.03 40 60 80 60 1 NA 20

LWA Kilns 0.5 0.015 30 60 80 1300 2.5 100 20

MACT Option 1b: Basic Incinerators ATF 0.2 0.015 30 60 80 25 1 100 20Option (includes common,protective D/F stnd andcommon Hg standard)


LWA Kilns ATF 0.2 0.015 30 60 80 1300 2.5 100 20

MACT Option 1c: Basic Incinerators ATF 0.2 0.015 ATF 5 60 80 25 1 100 20Option (includes common,protective standards forboth D/F and Hg)


LWA Kilns ATF 0.2 0.015 ATF 5 60 80 1300 2.5 100 20

MACT Option 2a: Basic Incinerators ATF 0.2 0.015 ATF 5 60 80 25 1 ATF 50 ATF 5Option 1c + ImprovedCombustion Cement Kilns ATF 0.2 0.03 ATF 5 60 80 60 1 ATF 50 ATF 5


MACT Option 2b: Basic Incinerators ATF 0.2 0.015 ATF 5 60 80 25 1 ATF 50 ATF 5Option 1c + ImprovedCombustion (W/CKexemption)



MACT Option 3: Basic Incinerators ATF 0.2 0.015 ATF 5 60 80 25 1 100 20Option 1c + CommonStandards Cement Kilns ATF 0.2 ATF 0.015 ATF 5 60 80 60 1 NA 20

LWA Kilns ATF 0.2 0.015 ATF 5 60 80 ATF 60 ATF 1 100 20

MACT Option 4: Basic Incinerators ATF 0.2 0.015 ATF 5 60 80 25 1 ATF 50 ATF 5Option 1c + CommonStandards + ImprovedCombustion

Cement Kilns ATF 0.2 ATF 0.015 ATF 5 60 80 60 1 ATF 50 ATF 5

LWA Kilns ATF 0.2 0.015 ATF 5 60 80 ATF 60 ATF 1 ATF 50 ATF 5

Nanograms per dry standard cubic meter of flue gas flow, a standardized volumetric measure of air3

releases.

Option 1a was eliminated from consideration after revisions to floor levels were made.4


! Floor Levels: These levels are based on current performance at existing facilities.

! Option 1: This option introduces above-the-floor (ATF) standards for dioxin, PMand mercury. Option 1b increases the dioxin control level to 0.2 ng/dscm while3

also introducing a more stringent mercury control level for cement kilns. Option1c introduces a common, protective mercury control level for all facilities (5ug/dscm).4

! Option 2: This option builds on Option 1c by adding improved combustionrequirements, as indicated by CO and HC emissions. Option 2b differs from 2aonly in that cement kilns are regulated less stringently (an HC control level of 20ppmv and no CO requirement).

! Option 3: This option builds on Option 1c by introducing a common PM standardwhere cement kilns must control at the same level as other facilities (0.015gr/dscf), a common HCl standard (60 ppmv) for cement kilns and lightweightaggregate kilns, and a common Cl standard (1 ppmv) for all facilities.2

! Option 4: This option is the same as Option 3 but also includes the improvedcombustion requirements found in Option 2a.

These regulatory variations imply significantly different costs and benefits. We give closest attention tothe impacts of the floor and Option 1b requirements, the regulatory alternatives that EPA viewed as mostlikely at the time this report was produced.

In addition to the regulatory options governing existing waste combustion facilities, this rule alsoproposes MACT standards for new sources. Exhibit 1-2 presents the specific standards for each HAP. Asshown, the basic option includes standards commensurate with the more stringent options for existingsources, as well as even more stringent semi-volatile metals and HCl requirements. Option 2 introducescommon CO standards for all sources as well as a more stringent low-volatile metals standard. Finally,Option 3 is equivalent to Option 1, except for more stringent dioxin and mercury requirements.


Exhibit 1-2

REGULATORY OPTIONSFOR NEW SOURCES


Option 1: Basic Option Incinerators 0.2 0.01 30 40 80 25 1 50 5

Cement Kilns 0.2 0.03 30 40 80 25 1 NA 20

LWA Kilns 0.2 0.01 30 40 80 25 1 50 5

Option 2: Basic Option Incinerators 0.2 0.01 30 40 30 25 1 50 5+ Common Standards +Stringent LVM Cement Kilns 0.2 0.015 30 40 30 25 1 50 5

LWA Kilns 0.2 0.01 30 40 30 25 1 50 5

Option 3: Basic Option Incinerators 0.1 0.01 5 40 80 25 1 50 5with improved D/F andHg Cement Kilns 0.1 0.03 5 40 80 25 1 NA 20

LWA Kilns 0.1 0.01 5 40 80 25 1 50 5


1.3 ORGANIZATION OF THIS REPORT

The remainder of this report consists of six chapters, plus appendices:

Chapter 2: Presents background information on the combustion sectors affected bythe technical standards rule, examining current waste burning and markettrends as well as current regulations and emissions of relevant HAPs.

Chapter 3: Explains the model plants methodology used to develop engineeringcompliance costs and compiles these costs across the universe of regulatedfacilities to arrive at upper bound estimates of national compliance costs.

Chapter 4: Presents our economic impact analysis. Using a series of screeninganalyses, we examine the differential impact that compliance costs arelikely to have on the various combustion sectors. In addition, we evaluatethe likelihood of facility closures if facilities are unable to cover newcompliance costs.

Chapter 5: Evaluates the cost-effectiveness of the rule by presenting information onexpenditures per unit reduction of relevant HAPs. In addition, we assesspotential health benefits of the rule.

Chapter 6: Presents the regulatory flexibility analysis for the technical standards rule,examining the potential for impacts on small business.

Chapter 7: Assesses the environmental justice implications of the rule, examining thecurrent impact of combustion facilities on various demographic andsocioeconomic subgroups located near combustion facilities.

Appendix A: Provides core data inputs to the combustion cost model such as prices.

Appendix B: Presents baseline costs of hazardous waste combustion for each regulatedsector.

Appendix C: Describes waste minimization options for wastes currently combusted.

Appendix D: Discusses waste management alternatives for wastes currently combusted.

Appendix E: Detailed Model Output for the floor and MACT Option 1B.

Appendix F: Detailed data on small business impacts to support the regulatoryflexibility analysis.

Appendix G: Presents market impacts assuming no price increase (zero percent price pass-through scenario).


OVERVIEW OF CURRENT COMBUSTIONPRACTICES AND MARKETS CHAPTER 2_______________________________________________________________________________________

The revised technical standards evaluated in this regulatory analysis address emissions from allfacilities burning hazardous waste except on-site boilers. The effect of the rule will not be the same in allcombustion sectors. As context for understanding how the rule will affect the industry, this chapterprovides a general overview of the combustion industry and the interplay between various sectors.Understanding the changes occurring within the combustion industry, even in the absence of the proposed MACT standards, provides important context for assessing the incremental impact of the rulein Chapter 4.

We first present an overview of the services provided by a combustion unit and the factors thatunderlie the demand for these services. Next, we describe the facilities that combust hazardous waste,including information on the current market structure and the quantity and characteristics of waste burned.We then provide background information on emissions from combustion devices and the installed base ofair pollution control devices (APCDs). Finally, we explore the current market performance of the variouscombustion sectors and evaluate some of the factors behind the differential performance of the commercialincinerator and the commercial boiler and industrial furnace (BIF) sectors. Factors examined includestructural advantages, differences in the regulatory baseline between sectors, and differences in the airpollution control devices currently in operation and in the pollutants released.

2.1 OVERVIEW OF SERVICES PROVIDED BY THEHAZARDOUS WASTE COMBUSTION SECTOR

Hazardous waste combustors provide three principal services to their customers when they burnwastes: regulatory compliance, liability reduction, and, in some cases, energy and raw material recovery.The demand for these services is driven by regulatory requirements, liability concerns, and economics.Regulatory forces affect the demand for combustion both by mandating certain hazardous waste treatmentstandards and by establishing requirements for the combustion units that affect the cost of combustion.Liability concerns of waste generators drive combustion demand because combustion destroys organicwastes, greatly reducing the risk of future environmental problems from the waste material. Finally,

Note that some, albeit much reduced, liability exposure remains in the form of residual incinerator1

ash that must be disposed of in a hazardous waste landfill. With some cement kilns and LWAKs, eventhis problem is minimized because much of the combustion residuals are integrated into the product.

Cary Perket and Jon Hanke, "The Roots of Overcapacity in the Incineration Sector," EI Digest,2

February 1995, p. 25.

Robert Graff and Thomas Walker, "Factors that Require, Encourage, or Promote Combustion of3

Hazardous Waste," memorandum to Walter Walsh, Office of Policy Analysis, U.S. EPA, prepared byIndustrial Economics, November 11, 1993, p. 12.

Graff and Walker, op. cit., p. 12; United States Environmental Protection Agency, Office of Solid4

Waste and Emergency Response, National Biennial RCRA Hazardous Waste Report (Based on 1989 Data),February 1993.


economic forces may encourage combustion where alternative management options are more expensive.Within the combustion segment, these same economic forces underlie the preference for energy and rawmaterial recovery over simple waste destruction as a way to reduce both combustion cost and potentialliability.1

2.1.1 Regulatory Requirements Encouraging Combustion

While industry began incinerating some of their hazardous wastes as early as the late 1950s, thecurrent market for hazardous waste combustion was essentially created by EPA regulation of hazardouswaste disposal. Two major regulatory forces directly encouraging combustion have been the land-ban2

restrictions under the Hazardous and Solid Waste Amendments (HSWA) of 1984 and clean-up agreementsfor Superfund sites called "Records of Decision (RODs)."3

Prior to the promulgation of EPA's Land Disposal Restrictions (LDRs), hazardous waste generatorswere free to send untreated wastes directly to landfills. The LDRs mandated alternative treatment for thewastes, known as Best Demonstrated Available Technologies (BDATs). Quite often, combustion was thestipulated BDAT. An assessment of the 1989 Biennial Report estimated that of the 24 waste codes(including mixtures) that comprise 80 percent of the waste reported combusted, 22 contained wastes withBDAT requiring some form of combustion. Pending LDRs could further increase the quantity of wastes4

flowing into the combustion sector.

The Land Disposal Restrictions have also influenced hazardous waste management under theComprehensive Environmental Reclamation, Compensation, and Liability Act (CERCLA). The RODs setout the cleanup plan for contaminated sites under CERCLA. A key attribute of the RODs is the choice ofremediation technology. Because even soil redeposited at Superfund sites is subject to the LDRs,incineration is often a technology chosen during remediation. While remediation efforts contribute a

Remediation share of combusted waste data are from Hugh Holman and D. Cotton Swindell,5

"Environmental Research: Chemical Waste Management and Rollins Environmental Services," AlexBrown & Sons, Inc., April 5, 1993, p. 8. Based on data in the 1991 BRS, about 15 percent of totalcombusted wastes were remediation wastes, treatment residuals, or other "one-time" wastes. See U.S.EPA, Setting Priorities for Hazardous Waste Minimization, July 1994, p. 2-12.

Graff and Walker, op. cit., p. 10.6

Graff and Walker, op. cit., p.15-16. This non-hazardous waste helps combustion units cover their fixed7

costs of operation, as well as competes for limited combustion capacity.

For larger wastes streams, however, waste segregation can often lead to large cost savings because less8

toxic fractions may be handled less expensively. See Michael Hoffman, "Waste Minimization VersusSegmentation," EI Digest, January 1994, pp. 45-46.


minority of the wastes managed by combustion (between 5 and 20 percent, depending on the incinerator),combustion has been used frequently on remediation projects. Between 1982 and 1991, incineration was5

the source control remedy selected most often (in 28 percent of RODs issued). 6

A number of regulations govern allowable emissions from combustion units and the processes bywhich residuals must be managed. By increasing or decreasing the cost of combustion, these regulationscan affect industry demand for combustion services to some degree. These regulations, which can alsoaffect the relative costs of different combustion options, are addressed in more detail in Section 2.4.2.1:Regulatory Advantages.

2.1.2 Liability Concerns

Liability concerns are mostly, but not entirely, driven by regulation as well. CERCLA created aliability system in which a generator who had shipped waste to a licensed disposal site could be held liablefor up to the entire cost to clean the site if environmental damages arose. With such large potential costs,generators found combustion's ability to destroy the wastes rather than simply dispose of them extremelyattractive.

Fears of product liability exposure through the courts have also increased demand for combustion.The Hazardous Waste Treatment Council estimates that 15 to 30 percent of waste handled by destructiveincineration is not classified as hazardous by any agency. Incinerators are often used to destroy off-7

specification or expired products that a manufacturer wants to ensure do not enter the marketplace.

2.1.3 Economic Forces Encouraging Combustion

Economic forces can encourage combustion over alternative treatment in two ways. First, becausecombustion can be used to treat a wide variety of waste streams, it is an attractive option for managerswishing to treat much of their waste on-site. In addition to ensuring control of the waste through itsdestruction, combustion may be cheaper than segregating and managing multiple small waste streams indifferent ways.8


Second, in a few market segments combustion is actually less expensive than alternative treatmentoptions, even in the absence of liability concerns. Presently, for example, some spent solvents arecombusted because combustion costs less than solvent reclamation. Due to the present low price ofpetroleum, the abundance of petrochemical manufacturing capital, and high compliance costs for solventrecyclers, reclaiming spent solvents has not been as cost-effective as paying facilities to combust the waste.

In addition to encouraging combustion over alternative treatment, economic forces play an evenstronger role in determining the competitive balance between sectors of the combustion market. Energyrecovery and residual reuse are currently possible in cement kilns and LWAKs, but not in commercialincinerators. This gives the kilns a cost advantage.

These economic forces are constantly changing. Some pending regulations are likely to increasethe flow of certain wastes to the combustion sector, while many others could change the relative costs ofcombustion for different market segments. Superfund liability reform currently under discussion inCongress could alter the importance of liability-reduction as a driver of combustion demand as well.Finally, technological changes in waste processing and product manufacturing influence the economicdrivers. The impact of these potential changes on market dynamics is addressed in Section 2.3.

2.2 HAZARDOUS WASTE COMBUSTION FACILITIES

Hazardous waste is combusted at a variety of different facility types. Two particularly instructiveways of classifying these facilities are by the type of unit or by whether the facility is commercial or non-commercial. Both factors influence the economics of waste burning at any given site. Unit types includeincinerators and boilers/industrial furnaces (BIFs). Incinerators traditionally burned wastes to destroy toxiccharacteristics, although some now recover a portion of the energy contained in the wastes. BIFs are unitsused to generate heat and/or power, or to manufacture products (like cement) in an industrial process.While BIFs traditionally burned conventional fuels like coal and oil, many have been modified to burnhazardous wastes as well.

Both incinerators and BIFs may either be commercial or non-commercial. Non-commercialfacilities are usually located at the generator's production site, and are referred to as "on-site" incineratorsor boilers. Commercial facilities generally receive wastes from generators in the surrounding region. Themain commercial combustion segments are commercial incinerators, cement kilns, and lightweightaggregate kilns (LWAKs). The commercial/non-commercial division is not always clear-cut, however;a few on-site facilities do accept some waste commercially even though most of the waste burnedoriginates on-site.

Companies that generate large quantities of waste usually choose to combust the waste themselves.On-site combustion offers a number of benefits to generators:

! Profits otherwise paid to commercial facilities are retained by the generator.

! The generator is somewhat insulated from price fluctuations in the commercialtreatment sector.

! Waste transport costs may be avoided all together.

In addition to energy recovery, cement kilns and lightweight aggregate kilns can incorporate a9

portion of the residual ash from combustion (of both hazardous and non-hazardous fuels) in theirproducts.

Phil Retallick, Rollins Environmental Services, personal communication, September 13, 1994.10

During 1991 and 1992, only eight percent of the energy used at kilns came from waste fuel. The11

hazardous waste fuel fraction was even smaller. See Portland Cement Association, U.S. Cement IndustryFact Sheet Twelfth Edition, June 1994, p. 17.


! Generators are able to control the entire treatment process, avoiding possibleliability risks.

! If wastes can be burned in an existing boiler or industrial furnace, the incrementalcost of waste burning can be very low. In such cases, the generator may alsoreduce energy costs somewhat by recovering the heat-content of the hazardouswastes.

For facilities that generate small to medium quantities of waste and do not already have an incinerator,paying a commercial facility to burn the waste is usually more cost effective than constructing andmaintaining an on-site incinerator.

Commercial and non-commercial combustion facilities either burn waste purely for destructivepurposes or for both destructive and energy recovery purposes. Incinerators, both commercial and non-9

commercial, burn waste primarily for destructive purposes. While some incinerators do use the cleanerliquid solvent streams to fuel their afterburners, energy recovery is not an integral part of their production10

process. In contrast, cement kilns, LWAKs, and generators with boilers are able to recover the energy inhazardous waste. For facilities like cement kilns with high energy requirements, the energy value of thewastes is an added benefit. However, even in the kilns, the waste destruction has a higher financial valuethan the recovered energy (see Appendix A). Furthermore, because of limitations on the amount ofhazardous waste that can be burned without affecting product quality, the majority of energy needed toproduce cement and lightweight aggregate is still derived from conventional fuels. 11

Exhibit 2-1 illustrates the current structure of the combustion market in terms of capacitydistribution across sectors and average permitted units per facility. With the exception of on-siteincinerators and on-site boilers (not characterized), a relatively small number of facilities combusthazardous waste. Both cement kilns and lightweight aggregate kilns tend to have more permitted wasteburning units per facility than do incinerators, and both use the energy content of the wastes to fuel theirproduction process.

See Daphne McMurrer, Bob Black and Tom Walker, "Memorandum: The Processing and Use of12

Waste Fuels," prepared by Industrial Economics for Lisa Harris, Office of Solid Waste, U.S. EPA,December 13, 1994.

Jeffrey Smith, "Fuel Blenders 1994," El Digest, September 1994, p. 20.13

Chris Goebel, National Association of Chemical Recyclers, personal communication, August 11,14

1994.


In addition to the actual combustion units, the combustion market contains a group ofintermediaries that move waste between the generator and the combustor. Waste brokers arrange themovement of wastes from the generator to the combustor without additional processing. In contrast, fuelblenders will generally collect waste from a number of generators and process it to meet the requirementsof their customers in the combustion sector. As of September 1994, there were approximately 74 active12

fuel blenders. The National Association of Chemical Recyclers (NACR) estimates that 40 percent of the13

waste received by its membership is recycled, while 60 percent goes to fuel blending. Of the fraction usedin fuel blending, less than 10 percent becomes residual waste that cannot be used as a fuel.14

Exhibit 2-1

NUMBER AND SECTOR OF COMBUSTION FACILITIES

Type of Combustion of Facilitie Units/Facility Commercial FunctionDevice Units s

Estimated Number Average Waste Commercial/Number of Burning Non-

Cement kilns 50 28 1.8 Commercial Energy recoveryand destruction

Lightweight aggregate kilns 15 7 2.2 Commercial Energy recoveryand destruction

Commercial incinerators 34 30 1.1 Commercial Primarilydestruction

On-site incinerators 184 157 1.2 Non-Commercial Primarilydestruction

Notes and Sources: (1) All units permitted to burn hazardous wastes have been included; however, some permitted units are no longer actively burning

wastes. (2) IEc estimate.(3) U.S. EPA, List of Permitted or Interim Status Hazardous Waste Incinerators and Hazardous Waste BIFs, November 1, 1994.

The wastes used as fuels are blended to meet customer requirements for energy content, viscosity,and acceptable concentrations of hazardous constituents. A consistent energy content is important both forBIFs and for incinerators. For BIFs, the waste fuels replace conventional fuels in a production process withspecific energy requirements. For incinerators, a variable thermal loading can reduce efficiency andpotentially damage the combustion unit. Viscosity affects the ability to pump wastes into the combustionunit in a uniform manner. Criteria for hazardous constituent concentrations are important both forcontrolling emissions and for protecting the stability of the production process and quality of the product(in the case of cement kilns and LWAKs). Fuel blenders have had a large impact on hazardous wastecombustion markets. This impact is described in greater detail below.

Jon Hanke, "Hazardous Waste Incineration 1994," EI Digest, June 1994, p. 19; Jeffrey Smith,15

"Industrial Furnaces 1994," EI Digest, October 1994, p. 23; Christine Seidel, "Another Look at the EPA's1991 Biennial Report," EI Digest, March 1995, p. 5.

Hanke, op. cit., June 1994, p. 19; Smith, op. cit., October 1994, p. 20.16

The final rule was published on February 21, 1991 (56 FR 7134).17

The 1994 weighted average heat value of fuels supplied to kilns by fuel blenders in the National18

Association of Chemical Recyclers (NACR) was 12,073 Btu/lb., with a minimum value of 8,800 Btu/lb.and a maximum of 14,000 Btu/lb. See NACR, NACR Waste Processing Survey, August 1994, question 1.Values vary by type of waste; see Appendix B for heat content values used in the EPA cost model.

Average heat content of waste at medium and large commercial rotary kiln incinerators from EER19

combustion database. See Appendix B.

Even wastes with a heat content greater than 5,000 Btu/lb. will often need to be blended with20

higher Btu wastes to keep the fuel feed to the kiln at an appropriate level.


2.3 THE QUANTITY AND CHARACTERISTICS OF COMBUSTED HAZARDOUS WASTE

Over the years, the quantity of waste combusted for destruction and energy recovery has steadilyincreased. Between 1989 and 1991, the tonnage combusted grew by 43 percent. Further increases wereapparent in the commercial sector between 1991 and 1993, although similar data for the on-site sector werenot available. 15

The overall growth of the market somewhat masks a slight increase in market share for commercialBIFs between 1991 and 1993. In 1991, an almost equal quantity of waste was burned by incinerators andthose facilities that burn waste for energy recovery. As shown in Exhibit 2-2, commercial and non-commercial incinerators burned 1.9 million tons with most of this occurring in on-site incinerators.Commercial BIFs burned more than 1.1 million tons, or about 62 percent of commercially combustedwastes. By 1993, commercial BIFs had captured 66 percent of the commercial combustion market, burning1.2 million tons compared to the 0.6 million burned in commercial incinerators. Although on-site BIFs16

are not included in this proposed MACT, we have included them in Exhibit 2-2 because wastes couldconceivably flow to them once the other sectors are regulated more stringently.

The characteristics of waste used for fuel differ from those of wastes sent to incineration, althoughfuel blenders have blurred the lines somewhat. The waste forms burned for energy recovery appear to bethose wastes most suitable for use as fuels (e.g., solvents and organic liquids). These waste forms are likelyto have a high heat content, be easy to pump, and result in a relatively small amount of solid residues. Thewastes managed in incinerators include those typically used as fuels as well as those wastes not amenableto energy recovery (e.g., lower Btu sludges and contaminated soil).

Under the Boiler and Industrial Furnace (BIF) rule, the waste burned for energy recovery must17

have a minimum heat value of 5,000 Btu/lb. In practice, the waste burned by cement kilns has an averageheat value of 12,000 Btu/lb, whereas incinerated waste typically has an average heat value of only 6,70018

Btu/lb. To make waste with a low heat content more amenable to commercial energy recovery, hazardous19

waste with a high Btu value is frequently mixed or "blended" with waste with a lower heat value. As20

Clyde Dempsey and Timothy Oppelt, "Incineration of Hazardous Waste: A Critical Review21

Update," Air & Waste, Volume 43, January 1993, pp. 47-48.


mentioned above, blending is also used to ensure that contaminants, such as metals, do not exceedallowable levels in fuels sent to combustion units.

2.4 MAJOR SOURCES OF COMBUSTED HAZARDOUS WASTES

Waste managed by combustion comes from a relatively narrow set of industries. As illustrated inExhibit 2-3, in 1991, the organic chemical industry (SIC 2869) alone generated nearly 40 percent of allcombusted hazardous waste (65 percent for SIC 28 overall). Based on available data, no other singleindustry generated more than 10 percent and, outside of SIC 28, no single industry generated more than 3percent. As shown in Exhibit 2-3, however, nearly one-quarter of all wastes combusted could not be tracedback to specific generator industries due to limitations in the BRS.

Much of the waste currently combusted is the result of petroleum refining and petrochemicalprocessing. While some producers in these industries are small (e.g., specialty chemical manufacturers),many are extremely large, vertically integrated, multinational firms.

2.5 BASELINE EMISSIONS AND EMISSION CONTROLS

This section presents information on baseline emissions from combustion devices as well as dataon the emission controls currently installed. Emissions and pollution control devices vary considerablyboth across and within combustion sectors.

2.5.1 Baseline Emissions

Baseline emissions data for hazardous air pollutants (HAPs) from combustion suggest thatemissions per combustion unit vary widely, and that there are different HAPs of concern in differentcombustion sectors. Emissions from BIFs arise from conventional fuel combustion as well as fromhazardous wastes. Cement kilns and LWAKs have a raw material feed that contributes to emissions aswell. While these factors make it difficult to attribute differences in emissions to hazardous waste burningalone, the data do highlight the potential problem areas for each sector.

The main constituents of concern are presented below, with the exception of products of incompleteproduction (PICs). Characterizing PICs is particularly difficult as the number of organic compoundspresent in stack emissions can measure in the hundreds and the type and quantity of PIC emissions varieseven under controlled operating conditions. Many PICs have not been characterized, and those thathave are often difficult to measure in stack emissions. As a result, many permit writers use carbonmonoxide (CO) and total hydrocarbon (THC) emissions as proxies for PIC formation. These values are21

presented in Exhibit 2-4 below.


Exhibit 2-3

INDUSTRIAL SECTORS GENERATING COMBUSTED WASTES*

Volume SIC Volume % of Cumulative %Rank SIC Code Description Code (tons) Volume of Volume

1 Industrial Organic Chemicals, N.E.C. 2869 1,147,907 37.79 37.79

2 Unknown Unknown 752,693 24.78 62.56

3 Pesticides and Agricultural Chemicals, N.E.C. 2879 287,214 9.45 72.02

4 Plastics Materials, Synthetic Resins, and 2821 172,634 5.68 77.70Nonvulcanizable Elastomers

5 Pharmaceutical Preparations 2834 114,390 3.77 81.47

6 Medicinal chemicals and Botanical Products 2833 98,137 3.23 84.70

7 Cyclic Organic Crudes and Intermediates, and Organic 2865 85,652 2.82 87.52Dyes and Pigments

8 Petroleum Refining 2911 78,700 2.59 90.11

9 Industrial Inorganic Chemicals, N.E.C. 2819 53,271 1.75 91.86

10 Refuse Systems 4953 31,083 1.02 92.88

11 Business Services, N.E.C. 7389 24,705 0.81 93.70

12 Photographic Equipment and Supplies 3861 21,642 0.71 94.41

13 Chemicals and Chemical Preparations, N.E.C. 2899 20,065 0.66 95.07

14 Nonclassifiable Establishments 9999 17,370 0.57 95.64

15 Synthetic Rubber (Vulcanizable Elastomers) 2822 12,289 0.40 96.05

16 Glass Containers 3221 9,038 0.30 96.34

17 Electric Services 4911 7,579 0.25 96.59

18 Paints, Varnishes, Lacquers, Enamels, and Allied 2851 5,788 0.19 96.78Products

19 Chemicals and Allied Products, N.E.C. 5169 5,559 0.18 96.97

20 Services, N.E.C. 8999 4,706 0.15 97.12

21 Manmade Organic Fibers, Except Cellulosic 2824 4,584 0.15 97.27

22 Manufacturing Industries, N.E.C. 3999 4,454 0.15 97.42

23 Wood Household Furniture, Upholstered 2512 4,341 0.14 97.56

24 Ammunition, Except for Small Arms 3483 3,658 0.12 97.68

25 National Security 9711 3,603 0.12 97.80

All Other SIC Codes 66,825 2.20 100.00

Total 3,037,866 100.00

* Excludes non-routinely-generated wastes and secondary wastes. Includes waste quantities burned in on-site boilers that are notsubject to this MACT.

Source: 1991 BRS Data in "Setting Priorities for Hazardous Waste Minimization," EPA, 1994.

Because emissions data were not subjected to rigorous statistical comparisons, differences in22

emissions across sectors should be used cautiously.


Using data on average emissions collected by EPA at a sample of combustion facilities, cementkilns appear to have the highest levels of emissions for all pollutants except low volatile metals (LVMs)and hydrochloric acid (HCl). Both total and per unit emissions of carbon monoxide (CO), mercury (Hg),22

particulate matter (PM), dioxin/furan toxic equivalents (TEQ), and THC appear considerably higher thanemissions levels for the incinerator and LWAK sectors. Incinerators appear to have the highest emissionsof LVMs, while LWAKs appear to emit the most HCl in total and per unit.

Exhibit 2-4

BASELINE NATIONAL EMISSIONS OF HAPs FROM COMBUSTION UNITS(pounds/year)

Cement Kilns Incinerators LWAKs

Emissions Unit Emissions Unit Emissions UnitAvg. Emissions/ Avg. Emissions/ Avg. Emissions/

Chlorine 421,277 8,426 1,957,156 8,978 58,615 3,908

CO 155,319,149 3,106,383 31,198,222 143,111 1,615,385 107,692

HCl 5,478,723 109,574 2,451,289 11,244 6,726,923 448,462

LVM 7,447 149 60,556 278 480 32

Mercury 29,681 594 10,173 47 696 46

PM 9,297,872 185,957 4,408,444 20,222 101,885 6,792

SVM 67,447 1,349 116,267 533 1,488 99

TEQ 2.07 0.041 0.187 0.0009 0.0002 0.00002

THC 10,393,617 207,872 529,013 2,427 97,385 6,492

Source: "Table 1. Baseline, floor, and above the floor (ATF) options national yearly estimated emissions," preparedby Energy and Environmental Research Corporation for Waste Management Division, U.S. EPA, April 21,1995. Data were rescaled to estimated units in combustion universe by IEc.

U.S. EPA, Combustion Emissions Technical Document (CETRED), May, 1994.23

These same data suggest that over half of the on-site boilers currently have no installed APCDs. As24

boiler emissions are not affected by this proposed rule, the lack of pollution controls may make the on-siteboilers more attractive economically.

Midpoint values from industry survey data presented in ICF Incorporated, 1990 Survey of Selected25

Firms in the Hazardous Waste Management Industry, July 1992, p. 2-5. Prepared for the U.S.Environmental Protection Agency, Office of Policy Analysis.


2.5.2 Baseline Pollution Control

Combustion facilities release a number of constituents of concern when they burn hazardouswastes. These constituents include particulate matter, metals, dioxin, mercury, chlorine and products ofincomplete combustion. The proposed regulation is designed to control risks from emissions of theseconstituents. Although nearly all facilities have installed some air pollution control devices, there aredistinct differences in the types of controls installed by various types of combustion facilities.

Exhibit 2-5 lists which constituents of concern are controlled by specific APCDs, as well as theprevalence of those APCDs by facility type. Data on APCDs are from EPA's Combustion EmissionsTechnical Resources Document (CETRED). Although CETRED is not all-inclusive in its characterization23

of installed APCDs, it does provide a useful overview of installed pollution controls. The types and24

efficiencies of controls currently installed has a large impact on the costs specific facilities will face undernew emission standards. Some of the interesting points that can be seen include:

! Only one facility currently uses carbon injection, a control technology which underthe proposed rule will frequently be required for dioxin/mercury controls.

! Although all cement kilns currently have some form of particulate controlinstalled, our model plants analysis (detailed in Chapter 3) shows that many kilnswill need to install a new fabric filter, either for PM control or as part of a carboninjection treatment train for mercury or dioxin controls.

! Lightweight aggregate kilns rely almost entirely on fabric filters for emissioncontrol.

! In contrast to the BIF sector, CETRED suggests that a small percentage ofincinerators currently operate with no emissions controls.

2.6 COMBUSTION MARKET PERFORMANCE

Throughout much of the 1980s, hazardous waste combustors enjoyed a strong competitive position.In spite of their high capital costs, incinerators were extremely profitable. EPA regulations requiringcombustion greatly expanded the waste tonnage requiring treatment. Federal permitting rules, as well asfierce local opposition to incinerator siting, constrained the entry of new combustion units. As a result,combustion prices rose steadily, reaching nearly $640/ton for clean high-Btu liquids and $1,680/ton forsludges and solids in 1987. Profits were equally high. For example, after-tax profits earned by Rollins25

Environmental Services, a firm operating primarily in the incineration sector, peaked at 16.4 percent that


Exhibit 2-5

BASELINE APCDS BY COMBUSTION SECTOR

Number (Percentage) of Sample UnitsCurrently Using Device

Control Device (1) Controlled Kilns Incinerators Incinerators KilnsEmissions Cement Commercial On-Site Aggregate

Lightweight

Fabric Filters Particulate 10 (29%) 10 (53%) 8 (16%) 8 (67%)matter, metals

Wet or Dry Particulate 25 (71%) 2 (11%) 6 (12%)Electrostatic matterPrecipitators (ESPs)

Cyclones Large 2 (4%)particulates

Wet or Dry Particulate 14 (74%) 38 (75%) 1 (8%)Scrubbers matter, dioxin,

volatile metals,acid gases

Spray Dryers Particulate 2 (11%) 1 (2%) 1 (8%)matter, acidgases

No Control Devices N/A 1 (5%) 1 (6%)(2)

No Information on N/A 1 (5%) 6 (12%) 4 (33%)Controls

Number of Units in N/A 35 19 51 12Sample

Percentage of Total 69% 56% 28% 79%Units in Sample N/A

Notes:

(1) Some controls, such as afterburners used to control PICs and hydrocarbons, were not broken out as anAPCD in CETRED. Some units have multiple control devices installed.

(2) Includes units specifically identified in CETRED as having no APCDs.

(3) Model plant analysis uses a more detailed breakout of baseline APCDs.

Source: IEc analysis of U.S. EPA, Combustion Emissions Technical Resources Document (CETRED) ,May 1994, Appendices.

Wayne Nef, "Rollins Environmental Services," Value Line, June 24, 1994, p. 352.26

ICF 1992, op. cit., p. 2-12.27

Data on capacity utilization in the commercial incinerator sector during 1993 are from Hanke, June28

1994, op. cit., p. 22.

Jon Hanke, "Hazardous Waste Incineration 1995," EI Digest, May 1995, p. 33.29

ICF, op. cit., p. 2-15.30

Jeffrey Smith, "Industrial Furnaces 1993," EI Digest, September 1993, pp. 19-21.31

Smith, October 1994, op. cit., p. 20. Excludes idled facilities and Marine Shale Processors.32

Utilization including idled facilities in 1993 was under 64 percent.

Ibid.33


year. The high profits induced many firms to enter the permitting and siting process for new combustion26

units, despite the inevitable delays in obtaining the required operating permits.

Hazardous waste combustion markets have changed significantly since the 1980s. Today there issubstantial overcapacity in the markets for both liquids and solids combustion. This has led to fiercecompetition, declining prices, poor financial performance, numerous project cancellations, and somefacility closures. In addition, there has been some consolidation of combustion capacity. The adversemarket environment has affected commercial incinerators most significantly. This section describes thecurrent market situation in more detail and discusses the key factors influencing market performance.

2.6.1 Overcapacity and Effects on Poor Market Performance

The hazardous waste combustion industry is currently overbuilt. Capacity utilization rates incommercial incinerators have fallen from 85 percent in 1985 to 56 percent in 1993, although they27 28

recovered to 70 percent during 1994. Utilization of solids capacity was significantly higher (at 74.329

percent) than for liquids (at 52.4 percent) in 1991, although no comparable data are available for lateryears. Over this period, however, the increased ability of BIFs to manage solids suspended in liquids has30

reduced the importance of this distinction. Furthermore, steep pricing declines for solids combustionindicate that solids capacity is no longer scarce.

Capacity utilization of cement kiln hazardous waste combustion capability was at roughly the samelevel as commercial incinerators in 1991, with approximately 59 percent utilization (versus 60 percent forincinerators). Combined capacity utilization for cement kilns and lightweight aggregate kilns was about57 percent. Utilization levels in 1992 increased to roughly 81 percent for cement kilns, but declined to31

68 percent in 1993. Utilization levels in LWAKs were slightly higher during 1993, at over 70 percent.32 33

Ibid., p. 25.34

Perket and Hanke, op. cit., p. 29.35

Jon Hanke, "Hazardous Waste Incineration 1993," El Digest, May 1993, p. 14.36

Holman and Swindell, op. cit., p. 15.37

See Christine Seidel, "Mobile Thermal Treatment 1994," El Digest, December 1994, pp. 21-26.38


This overcapacity is the result of a number of factors, including:

! New Supply. The new supply comes both from new and expanded combustionunits. Although many projects have been canceled, some units begun prior to theprice declines of the past few years continue to come on-line. The elimination ofwaste processing bottlenecks (e.g., waste storage capacity) has also expanded thecapacity of some facilities already in operation.

! Increased Solids-Burning Capability in BIFs. Fuel blenders have improvedtheir ability to suspend solids in liquid wastes. This has greatly expanded theeffective solids burning capacity among BIFs that could previously only burnliquids, and driven down prices in this formerly high-profit segment. Between1989 and 1994, the average suspended solids content in waste fuel rose from 18.2percent to 22.2 percent. Industry anticipates that the solids content will reach up34

to 30 percent by the end of the decade. 35

! Waste Minimization Efforts. Industry efforts to minimize hazardous wastegeneration have reduced the quantity of wastes requiring treatment. This reductionhas been estimated at between 5 and 10 percent per year.36

! Substitution of Alternative Technologies in Remediation Market. Much of thefuture combustion demand for remediation of government-owned sites is likely tobe handled by on-site units. New alternative technologies, such as thermal37

desorbers, have further weakened demand.38

While EPA-promulgated land disposal restrictions (LDRs) during this time period increased wastequantities managed in the combustion sector, these increased quantities were insufficient to offset thefactors described above. In addition to factors affecting all combustion segments, there have been otherforces influencing the competitive balance between commercial incinerators and BIFs.

Incinerators can use some cleaner solvent streams to fuel their afterburners. However, while some39

broader energy recovery is done at European incinerators, it is unlikely to be done in the United States.When the heat recovery process runs hot gas through a heat exchanger, the temperature of the gas flowdrops, increasing the likelihood that chlorine PICs can re-form dioxins. This increases the dioxin emissionsfrom the stack. (Retallick, op. cit.)


2.6.2 Existing Market Advantages for BIFs

Industry overcapacity has led to intense competition in the combustion segment and has putpressure on the industry's less-efficient players. Due both to structural and regulatory advantages, BIFsappear to have had the edge recently in the current market place.

2.6.2.1 Structural Advantages

BIFs possess a number of structural advantages in the combustion of hazardous wastes that willremain regardless of federal regulatory actions. Foremost is the ability to use existing production capitalequipment to combust hazardous wastes and to utilize the energy content of the wastes in their productionprocess. The result is that the incremental cost to burn a ton of hazardous waste in a BIF is significantlylower than the cost to burn it in an incinerator.

! Energy Recovery. BIFs can make use of the heating value of hazardous wastefuels to offset purchases of virgin fuels that would otherwise be needed to achieverequired heating temperatures to a much greater extent than can incinerators. A39

commercial incinerator uses process heat to break down and destroy hazardousorganic wastes, while a cement kiln uses the heat both to break down wastes andto manufacture cement, a saleable end product.

! Shared Capital. Even in the absence of energy recovery advantages, cementkilns still enjoy an advantage based on their ability to produce a saleable product.A commercial incinerator must purchase all of its capital equipment to combusthazardous wastes and control emissions from the process. In contrast, a cementkiln purchases capital equipment to manufacture cement, and this equipment canalso be used to destroy hazardous wastes. While there are some incrementalcapital purchases required for a kiln to burn hazardous wastes, these are smallrelative to the overall cost of an incineration unit.

2.6.2.2 Regulatory Advantages

Under current regulations, kilns are not required to dispose of their ash in a Subtitle C hazardouswaste landfill. In addition, various industry segments have argued that the regulation of emissions is notconsistent across the industry. Differences in emissions by unit type and in air pollution control devices(APCDs) currently installed provide some insights into these claims. However, as is noted below, theregulatory baseline is continually changing, and advantages presented here are likely to change as well.

Data are for 1993. High-end estimate includes stabilization. Transport costs would be in addition to40

the prices quoted. See Jon Hanke, "Hazardous Waste Landfills 1994," EI Digest, April 1994, p. 30.

U.S. Environmental Protection Agency, Report to Congress on Cement Kiln Dust, 1993, p. 9-10.41

In 1995, EPA issued a regulatory determination stating that CKD management should be regulated.42

A recent industry proposal, less stringent than some options EPA has developed, which would place CKDin a monofill, has been developed as well. Even this approach would be more expensive than currentmanagement methods. See Bureau of National Affairs, "Cement Industry `Enforceable Agreement' WouldReplace Agency's Plan for Kiln Dust," Environment Reporter, March 31, 1995, p. 2371.

As of June 1995, for example, all waste burning cement kilns were operating under interim status. 43

(Karen Randolph, U.S. EPA, personal communication, June 13, 1995.)


Ash Disposal

Ash from hazardous waste incinerators is itself considered a hazardous waste. The material mustbe disposed of in a permitted hazardous waste landfill at a cost of $96 to $158 per ton. In comparison,40

ash from cement kilns or LWAKs can either be integrated into the facility's products, sold, or dumped on-site as a non-hazardous material at a cost of slightly over $3/ton.41

EPA regulatory initiatives are likely to change this balance within a few years. Future regulationof cement kiln dust (CKD), as ash from cement production is called, will likely increase the cost ofmanaging residuals at kilns that combust hazardous wastes. (However, to the extent that waste-burning42

and non-waste-burning kilns face the same CKD management costs, cement markets rather than waste-burning markets are likely to be affected.) Similarly, the Hazardous Waste Identification Rule (HWIR),scheduled for release in 1996, may allow some treated hazardous wastes to exit the hazardous wasteregulatory system. Thus, if ash from incinerators no longer exhibited hazardous properties, it could bemanaged as non-hazardous, reducing the cost to incinerators.

Other Differential Regulations

Differences in the requirements for fully permitted facilities can create advantages for one sectorover another. In addition, interim status under the BIF rule can create temporary benefits for BIFs thatdisappear once a unit is fully permitted. Representatives from each industry claim that their facility type43

is more stringently regulated than the other, and thus subject to higher costs. In addition to differences inthe disposal requirements for combustion residuals, already discussed above, commercial incineratorindustry representatives claim that:

Comment is from Richard Fortuna, "The Combustion Strategy and the HWTC," EI Digest, December44

1993, p. 8. A petition introduced to EPA by the Cement Kiln Recycling Coalition in January 1994 calledfor metals standards five to ten times more stringent than the emissions levels in the current BIF rules.Representatives of the commercial incinerator industry note this act as an admission that controls in the BIFregulations are too lax. (See Bureau of National Affairs, op. cit., p. 1645).

Fortuna, op. cit., p. 8. 45

See Cement Kiln Recycling Coalition, "The Combustion Strategy and the CKRC," EI Digest,46

December 1993, p. 12.

The incinerator regulations do not require metal emissions standards, but limit particulate matter47

emissions. Since low particulate matter emissions do not necessarily correspond with low toxic metalsemissions, opponents view the controls as inadequate. (Bureau of National Affairs, op. cit., p. 1645).

Hanke, June 1994, op. cit., p. 23.48


! BIFs have lax standards for metal emissions relative to commercial incinerators.44

! The destruction and removal efficiency (DRE) verification does not need to occurfor BIFs until a full permit is issued.45

Conversely, the BIF industry asserts that incinerators have an advantage under current regulations. Theyargue that BIF regulations are more stringent, even as implemented during interim status, than the "12-year-old requirements in Subpart O for commercial incinerators." For example, Subpart O regulations do not46

require extensive feed rate analysis on a continuous basis and do not establish metal-specific emissionlimits.47

The validity of these claims is difficult to gauge. Baseline emissions shown in Exhibit 2-4 suggestthat BIFs have higher average emissions of mercury and semi-volatile metals than do incinerators.Incinerators emit more low volatile metals. However, these data cannot be used to compare emissions perton of waste burned across sectors. Nor to they provide insights into the cost savings to any sectorattributable to higher emissions. The proposed MACT will likely eliminate any emissions-based costadvantages by ensuring that hazardous waste combustion in all regulated sectors is done in a mannerprotective of human health and the environment.

2.6.3 Market Performance Across Combustion Sectors

While the combustion sector overall has experienced declining prices, commercial incineratorshave been hit harder than BIFs. In the commercial incineration sector, average prices for liquid organicsfell by 10.4 percent between 1991 and 1993. Solids prices declined by 17.4 percent. Prices in the cement48

kiln sector remained mostly stable over this period, as measured by the prices that fuel blenders paid tocement kilns. Kilns continued to accept wastes for less money than incinerators, earning on average 65percent less on liquids and about 37 percent less on solids. This may be due in part to the kilns' lower costsand in part to the higher heat content of the waste streams they receive. Although prices began to

Perket and Hanke, op. cit., p. 30.49

Data on the prices charged to generators are difficult to compare exactly due to differences in the50

timing of data collection and a lack of information about the contaminant level in the waste streams. SeeHanke, June 1994, op. sit., p. 23 and Smith, September 1994, op. cit, p. 28.

Rollins (including combustion units recently purchased from Aptus) owned an estimated 25.5 percent51

of commercial incinerator practical capacity as of May 1995. See Hanke, May 1995, op. cit., p. 35.

Wayne Nef, "Rollins Environmental Services," Value Line, March 24, 1995, p. 350.52


strengthen in late 1994, analysts do not think that continued growth in waste volumes will be sufficient tosustain these gains.49

Prices charged to generators fell in both the commercial incineration and kiln sectors. Averageprices charged to generators seem to be quite similar whether the wastes are burned in incinerators orkilns. In contrast, prices paid to kilns by fuel blenders over this period remained constant, while prices50

paid to incinerators declined. This trend makes sense given the kilns' lower baseline cost for hazardouswaste combustion (see Appendix B), because kilns appear to be the pricing leaders. A number of othermarket indicators, including profits, returns to shareholders, and capacity expansions and closures highlightthe difficulty incinerators face relative to kilns.

2.6.4 Financial Performance

Financial performance indicators help contrast the condition of incinerators and cement kilns, butare subject to two caveats. First, financial data for Rollins Environmental Services is used as a proxy forthe entire commercial incinerator sector because data on other firms include substantial non-incineratorassets, and because Rollins is a large portion of the industry. Performance of incinerators owned by other51

firms may be somewhat different from Rollins, though we have no reason to believe these differences arelarge. Second, financial performance for cement kilns is heavily influenced by the cement markets.Nonetheless, the baseline costs of hazardous waste combustion in the kilns (detailed in Appendix B)suggest strong returns on waste burning.

Profitability

Examining financial returns for Rollins Environmental Services provides some insights into theeconomics of the incineration segment of the market because Rollins derives nearly 80 percent of revenuesfrom incineration. The firm's net profit margin peaked at 16.4 percent in 1987, and remained quite highuntil 1992. Between 1984 and 1992, the net profit margin averaged over 13 percent. This dropped to 5.5percent in 1993, and the firm lost money in 1994. Analysts expect net profits to increase to 3.8 percentduring 1995. 52

Thomas Mulle, "Cement and Aggregates," Value Line, January 20, 1995, p. 891.53

"Commercial Hazardous Waste Management: Financial Performance and Outlook for the Future,"54

annual report from the July/August 1985, 1987, 1988, 1991, and 1993 issues of The Hazardous WasteConsultant, op. cit.

Nef, March 24, 1995, op. cit., p. 350.55

Mulle, op. cit., p. 891.56


Profits in the cement industry are presently stronger than in the commercial incineration segment.From a minuscule 0.1 percent net profit margin in 1991, in the midst of the recession, margins recoveredto 2.8 percent in 1993 and are estimated at 4.9 percent in 1994. Net margins are projected to reach 6.7percent by 1995, substantially higher than projected levels for Rollins. 53

Despite the projected increase in profits for the cement industry, neither sector currently has astrong profit margin. It is important to note, however, that the returns on cement overall, a commodityitem, can mask much higher returns currently being earned on the combustion of hazardous wastes atcement plants. It is the return on hazardous waste burning (analyzed in Chapter 4) that will drive decisionsby plant operators on whether or not to expand waste handling capacities.

Returns to Shareholders

The return-on-equity ratio (ROE) measures the financial returns to investors in a firm or industry.As these returns fall, it becomes more difficult for firms to raise new funds in capital markets. ROE forthe hazardous waste management industry peaked at over 25 percent in 1984, dropping to nine percent in1989, and recovering to 16.3 percent in 1992. 54

Again, Rollins Environmental Services can serve as a more direct proxy for commercial hazardouswaste incinerators. Rollins' ROE between 1985 and 1988 was above 20 percent, a better performance thanthe environmental services sector overall. As incineration overcapacity became apparent, Rollins' ROEdeclined steadily to only 5.6 percent in 1993. After losses in 1994, the firm's ROE is expected to rise toonly 4.5 percent this year.55

Average returns to shareholders in the cement industry dropped from 8.6 percent in 1990 to only0.1 percent in 1991 as a result of the recession. Although ROE had recovered only to 6.8 percent in 1993,scarce capacity for cement is expected to increase shareholder returns sharply, reaching 13 percent in1995. This implies that the cement industry may be able to raise investment capital more readily than the56

commercial incineration sector over the next few years.

"1993 Outlook for Commercial Hazardous Waste Management Facilities: A Nationwide Perspective,"57

The Hazardous Waste Consultant, March/April 1993, p. 4.2.

"1994 Outlook for Commercial Hazardous Waste Management Facilities: A North American58

Perspective," The Hazardous Waste Consultant, March/April 1994, p. 4.4.

Seidel, op. cit., January 1995, p. 6.59

Smith, September 1993, op. cit., p. 18.60

Smith, October 1994, op. cit., p. 25.61

Ibid., pp. 19, 25.62

Ibid., p. 24.63

Hazardous Waste Consultant, March/April 1994, op. cit., p. 4.4.64

Smith, October 1994, op. cit., p. 23.65

Seidel, January 1995, op. cit., p. 6.66


2.6.5 Capacity Expansions and Cancellations

With industry overcapacity, falling returns, and new alternative technologies on the horizon, newincinerator projects have been canceled at a steady rate over the past five years. Twenty-five projects werecanceled between 1990 and 1992, more than in the previous seven years. During 1993, no new57

incinerators were proposed, and seven pending incinerators and two pending incinerator expansions wereabandoned. Three new incinerators plan to come on-line during the coming year despite the glut of58

capacity. However, these units mark the completion of earlier construction, and the initiation of other59

new projects seems unlikely at this time.

Cement kiln capacity to combust hazardous wastes, on the other hand, has grown and is expectedto increase at 10 to 20 percent per year for the next few years. Eight facilities (over one-quarter of the60

facilities now burning hazardous wastes) plan capacity increases as they move toward final part B permitstatus. While some new kilns will be brought on-line, improved load management (such as with waste61

storage tanks and better waste pumping systems) will be the source of much of the increased capacity.62

Despite some planned capacity increases, however, kilns have not been immune from the overallstagnation in combustion markets. While capacity to burn has continued to climb, the number of kilnsburning wastes is expected to decline due to the BIF regulations and to market conditions. During 1993,63

eight cement kiln projects were abandoned, and nine kilns had permit requests denied. The tonnage of64

waste burned in commercial BIFs actually declined slightly, from 1.20 million tons to 1.16 million tons,in 1993. During 1994, Southdown announced that its two waste burning kilns would exit that market.65 66

EPA's combustion strategy, the BIF rule, and increasing local opposition to hazardous waste combustionat the kilns all played a role.

Hanke, June 1994, op. cit., p. 24.67


2.7 SUMMARY

The proposed MACT standards would more stringently regulate emissions of hazardous airpollutants from commercial hazardous waste incinerators, on-site incinerators, waste-burning cement kilns,and lightweight aggregate kilns. BIFs burn the most waste commercially, holding a 31 percent share ofthe combustion market in 1991. Commercial incinerators had a 19 percent share, but this has been fallingin recent years. On-site incinerators were the largest combustion outlet, burning 32 percent of thecombusted wastes in 1991. The primary source of these wastes is the chemical industry (SIC 28), whichalone comprised approximately 65 percent of all regularly-generated primary waste combusted during1991. No other single industry generated more than three percent.

The magnitude of current emissions from combustion varies by HAP. For all HAPs other thanchlorine, low volatile metals, and HCl, cement kilns had the highest emissions per combustion unit. Formany HAPs, including CO, Hg, PM, TEQ, and THC, per unit emissions from cement kilns appear to beconsiderably higher than from other types of combustion units.

Although hazardous waste combustion markets have recovered slightly in the past year, substantialovercapacity exists in the commercial HW combustion section, a fact acknowledged by most industryparticipants. While waste generators will continue to demand compliance, liability reduction, and67

economic efficiency in managing their hazardous wastes, options other than traditional commercialincineration continue to expand. These include combustion in BIFs, other innovative technologies, andsource reduction. As a result, the combustion industry is unlikely to recover profitability levels experiencedin the late 1980s.

Future market performance will be determined by two key factors: combustion unit closures andchanges in the regulatory baseline. Even without additional regulation, closures will be necessary to bringcapacity utilizations up. Based on current market dynamics, structural advantages for BIFs suggest thatin the long run the bulk of closures will be in the commercial incinerator sector. However, someincinerators will retain market share because of their ability to combust wastes that BIFs either cannot ordo not wish to burn. Changing regulations will also affect the degree to which cement kilns retain their costadvantage in combustion markets. We examine the impact of this proposed rule, and other key regulatorychanges, on the structure of combustion markets in Chapter 4.


ENGINEERING COST ANALYSIS CHAPTER 3_______________________________________________________________________________________

Combustion facilities complying with the maximum achievable control technology (MACT)emissions standards will likely achieve the required emission reductions by installing pollution controldevices and instituting other operating measures. The purpose of this chapter is to review baselineestimates of the costs associated with these compliance actions. We first discuss the methodology used toestimate national compliance costs, including:

! the model plants approach used to assign pollution control measures to individualfacilities; and

! the cost aggregation methodology used to scale up the model plant costs to thenational level.

The results section of this chapter reviews the estimates of total and per unit compliance costsassuming all facilities in the industry install controls to comply with the regulations. As we explain inSection 3.1.2.1, these costs are likely to represent an upper bound estimate of the expected costs of theproposed rule. Our economic impact analysis (Chapter 4) presents a refined cost estimate that takes intoaccount units that stop burning hazardous waste and changes in waste flows that ultimately reduce theactual costs of regulation. In addition to the total cost estimates for existing waste burning facilities, wealso present per unit compliance costs associated with the proposed MACT standards for new sources.

3.1 METHODOLOGY

3.1.1 Model Plants Approach

As a starting point for assessing economic impacts, the Environmental Protection Agency (EPA)developed an analysis of the engineering costs associated with the MACT floor and above-the-floor (ATF)emission standards. Because of the diverse characteristics and large number of regulated combustion units,EPA applied a model plants approach whereby individual units are assigned to a model plant category onthe basis of current emissions and types of new air pollution control equipment required to comply withthe proposed rule. The model plant designation represents the set of pollution control measures that willbe necessary to meet the standards for each Hazardous Air Pollutant (HAP) in a given regulatory option.

Cost Estimates for Air Pollution Control Device (APCD) Requirements for Existing Facilities to Meet1

Proposed MACT Standards for the Floor and Above the Floor Options for Cement Kilns, Lightweigh tAggregate Kilns, and Incinerators, prepared for U. S. EPA, Office of Solid Waste, prepared by Energy andEnvironmental Research Corporation, April 1, 1995.

Combustion Emissions Technical Resource Document (CETRED), U.S. EPA, Solid Waste and2

Emergency Response, May 1994. Note that use of test burn data may bias the measurement of emissionsby including data from spike testing and other abnormal operating conditions.


From this assignment of pollution control devices, we can derive for each option the capital and operatingcosts that each modeled unit would incur. The proposed MACT options (shown in detail in Exhibit 1-1)differ from one another in the stringency of the emission limits.

A detailed discussion of the model plants methodology and results can be found in a separatedocument produced for EPA. We review the general elements of the approach below (also shown in1

Exhibit 3-1), including the model plants approach and the approach for assessing an alternative complianceoption whereby facilities reduce waste fed to the combustion unit. Finally, we review the limitations ofthe model plants approach.

3.1.1.1 Model Plant Cost Estimation Procedure

The first step in the model plants approach is to identify the emission rates of each HAP at eachexisting combustion unit. Emission rate data are based on test burn information compiled for EPA'sCombustion Emissions Technical Resource Document (CETRED). The analysis assigns a single emission2

rate for each HAP by taking each combustion unit's average over all test condition runs.

Once this emission rate is established, the second step in the analysis is to calculate the emissionreduction that is required to meet the MACT floor or ATF level at each combustion unit. In cases whereemissions data for a particular HAP at a particular unit are not available, the percentage emission reductionis assigned (0, 25, 50, or 75 percent) based on the distribution of emission reductions required for otherfacilities in the same combustion category. While the value for any one HAP may be incorrect, thepopulation of combustors as a whole will be fairly represented.

In the third step of the model plants analysis, the appropriate control devices are selected based onrequired emission reduction estimates. If the emissions reduction required for the HAP is modest and canbe achieved with devices that currently exist at the facility, the assigned control measure will involvechanging the design, operation, and maintenance (DOM) of the existing equipment. For example, a modestparticulate matter (PM) reduction may be achievable by optimizing the cleaning cycles and test procedureson an existing fabric filter system. If no existing device controls the HAP, or if a significant reduction isrequired, the model plants analysis assumes that a new device will be installed at the facility. Exhibit 3-2provides a brief overview of the categories of control devices included in the model plants analysis. Thespecific criteria driving the selection of control devices can be found in the model plant documentreferenced above.

At first glance, there appear to be more model plants than actual combustion units. In reality, this is3

not the case. Were each combustion unit modeled separately for each proposed MACT option, there wouldbe a total of nearly 2,000 permutations (283 combustion units x 7 MACT options = 1,981).


Exhibit 3-2

EMISSION CONTROL DEVICES ASSIGNED IN MODEL PLANTS ANALYSIS

HAP Devices Applied Comments

PM, Low-Volatile - Fabric filterMetals, Semi-VolatileMetals

HCl and Chlorine - Packed tower- Wet scrubber- Spray tower (LightweightAggregate Kilns (LWAKs))

Mercury - Carbon injection/carbon bed Carbon injection must be accompanied by adry PM control device.

Dioxin/Furan - Temperature control Temperature control applicable only at- Carbon injection units operating at higher temperatures.

Hydrocarbons and CO - Afterburner Must be accompanied by a water quenchsystem. For incinerators requiring limitedreductions, operating changes aresufficient.

Source: Energy and Environmental Research Corporation, Cost Estimates for APCD Requirements forExisting Facilities to Meet Proposed MACT Standards for the Floor and Above the Floor Optionsfor Cement Kilns, Light Weight Aggregate Kilns and Incinerators , Draft Report prepared for theOffice of Solid Waste, U.S. EPA, April 1, 1995.

The fourth step is to define model plants based on different combinations of required controlmeasures. The set of emission controls required by the units in the group defines that particular modelplant. For example, Model Plant Group 8 for cement kilns is defined as those kilns that must add carboninjection, a packed tower, and a fabric filter. Each model plant is further subdivided into sizeclassifications. Grouping units by size is important because the cost of APCDs can vary widely acrosslarge and small units. The average flue gas flow for each subgroup is used in the model plant assessment.For example, all cement kiln model plants are divided into large and small cement kilns; small kilns aredefined as having a gas flow of 147,000 actual cubic feet per minute (acfm) while large kilns have a flowof 370,000 acfm. The definition of these model plants is constant, i.e., it does not change under differentregulatory options. Instead, what changes is the distribution of combustion units to the model plantcategories. Changes in this distribution are summarized in Table 67 of the model plants report. Across allregulatory options, there are a total of 127 model plants: 30 for cement kilns, 19 for LWAKs, and 78 forincinerators. A single combustion unit can fall into new model plants under different MACT options. 3

Energy and Environmental Research Corporation (EER), April 1, 1995, op. cit.4


The fifth step in the model plants analysis involves assigning each of the existing combustion unitsto model plant groups on the basis of the required devices and DOM measures. The model plants reportprovides the assignment of individual combustion units to the model plants in the fourth table under eachregulatory option (e.g., Tables 4, 9, 13, 17, 21, etc.).

The final step in the model plants analysis is the assignment of engineering costs to each modelplant. EPA estimated both capital and operating costs for each air pollution control device usingspreadsheet models developed by EPA's Office of Air Quality Planning and Standards (OAQPS). OAQPSspreadsheet models exist for a number of the key air pollution control devices such as scrubbers, fabricfilters, and carbon injection units. OAQPS developed the models based on data from municipal wastecombustors, pollution control equipment vendors, and the OAQPS Control Cost Manual. Each modelrequires user-specified inputs on facility size (measured by flue gas flow), operating temperature, and otherparameters. The spreadsheet uses these inputs in a series of equations that provide total and annual capitalcosts, as well as annual operating and maintenance (O&M) costs.

In conducting the model plants analysis, EPA customized the OAQPS cost models for eachtechnology at each model plant. For example, the analysis specifies appropriate parameter values (e.g.,gas flow rate for the combustion unit) for a wet scrubber system that would be installed and operated tocontrol HCl emissions from a lightweight aggregate kiln. As mentioned, to account for differences in thesize of combustion units, the model plants analysis splits each model plant category into small, medium,and large facilities. Each combustion unit in the analysis is assigned to the size category on the basis offlue gas flow rate. This allows the capital and operating costs to be more accurately calibrated to the unitsin the model plant category. The compliance cost for any given combustion unit is the sum of the costsassociated with each of the control technologies used at that unit. The capital, operating, and totalannualized costs for each model plant are summarized in Appendix D of the model plants report.4

3.1.1.2 Continuous Emissions Monitoring Costs

EPA is considering whether to require continuous emissions monitoring (CEM) at all of thehazardous waste combustion facilities regulated by the proposed MACT standards. As the name implies,CEMs would allow regulators to track emissions from combustion facilities continuously. Emissions datacould either be transmitted from the facility to data receiving points at EPA and/or state agencies, or storedon-site and reviewed during an inspection. This would represent an alternative to the current systemwherein emissions are regulated on the basis of trial burn data gathered for permit applications andrenewals and routine measurement of operating parameters. CEMs would allow EPA to enforce theproposed MACT standards more closely and would ensure that violations of the standards do not occurbetween periodic emissions tests.

EPA has estimated the per-unit cost of implementing the CEM requirements. As shown in Exhibit3-3, these costs depend directly on the set of HAPs monitored. Under one option (Option 1), only carbonmonoxide and hydrocarbons would be monitored. The annualized cost for this monitoring is approximately$80,000 per combustion unit. Under the second option being considered, CEMs to monitor PM andmercury, as well as CO and hydrocarbons, would be required. As shown, the mercury monitoring addssignificantly to the overall cost of this option (about $40,000 per year). Total costs under Option 2 would

The method for estimating these costs can be found in "Revised MACT Compliance Costs,"5

memorandum prepared for Larry Denyer, EPA, prepared by Greg Kryder, Wyman Clark, EER, April 24,1995.


be about $130,000 per unit per year. A final option would have each facility monitor an extensive set ofHAPs, including CO, hydrocarbons, PM, HCl, Cl , Products of Incomplete Combustion (PICs), and2

mercury. The annualized cost of this option would be approximately $190,000 per unit, with the cost ofPIC monitoring being the greatest increment relative to Option 2.5

Exhibit 3-3

ESTIMATED PER-UNIT ANNUAL COST OFCONTINUOUS EMISSIONS MONITORING

($ thousands)

Small Large Small Med. LargeHAPs Monitored CK CK LWAK Incin. Incin. Incin.

Option 1: Baseline CEMSystem (CO, HC) $77.5 $80.5 $84.9 $80.4 $81.5 $84.2

Option 2: Baseline Systemwith PM and Hg CEM

$125.5 $128.4 $132.9 $127.0 $128.5 $131.7

PM Increment $7.6 $7.6 $7.6 $7.1 $7.3 $7.6

Hg Increment $40.4 $40.4 $40.4 $39.5 $39.7 $39.9

Option 3: Full CEM System(CO, HC, PM, HCl, Cl ,2

PICs, Hg) $187.5 $190.4 $192.7 $187.3 $188.8 $192.0

PIC and Cl2 Increment $47.7 $47.7 $45.5 $46.0 $46.2 $46.6

HCl Increment $14.5 $14.5 $14.5 $14.8 $14.8 $14.9

Source: "MACT Compliance Costs," memorandum prepared for Larry Denyer, EPA, preparedby Greg Kryder, Wyman Clark, EER, April 8, 1995; revised April 24, 1995.

For the purposes of assessing the economic impact of the proposed MACT standards, we haveassumed that Option 2 will be required. The annual costs of CEMs are added to the annual costs associatedwith pollution control measures in the model plants analysis to arrive at total annual cost per combustionunit.

Jon Hanke, "The Search For Continuous Emissions Monitors," EI Digest, March 1995, pp. 8-12.6

Installation of an Electrostatic Precipitator (ESP) field DOM is the exception; this change would7

require between three and eight weeks, depending on the combustor type.

"Shutdown Time Estimates," memorandum prepared for Bob Holloway, EPA, prepared by Greg8

Kryder, et al., Energy and Environmental Research Corporation, May 5, 1995.


We should note that these estimates of CEM costs are subject to significant uncertainty. CEMsystems for many of the HAPs are still under development and therefore difficult to cost out with precision.All required CEMs are expected to be readily available by the time the proposed rule is promulgated,however. 6

3.1.1.3 Other Compliance Costs

In addition to monitoring costs, EPA also anticipates that facilities will incur an incremental annualcost of approximately $5,000 for permit modification expenditures. This includes legal costs, engineeringcosts, and a trial burn plan. These costs are based on estimates developed for the Commercial Boiler andIndustrial Furnace (BIF) rule and will be revised based on an information collection request (ICR) currentlybeing conducted by EPA.

We also investigated the significance of shutdown costs associated with installation of the pollutioncontrol equipment assigned in the model plants analysis. Depending upon the model plant category,combustion facilities and their associated enterprises could incur costs during periods when the productionprocess must be discontinued to install controls. For example, if a cement plant must cease operating forseveral weeks to install and test new pollution control devices, revenues on both cement and waste burningcould be lost.

Examination of shutdown times suggests that virtually all of the installations could be coordinatedwith routine maintenance shutdowns. We assessed shutdown times based on a vendor survey of periods7

required for device installation, Portland Cement Association comments on the OAQPS cost models, andengineering judgment. We assumed that all retrofits could be made simultaneously during a single facilityshutdown. In addition, we assume that combustion units normally would be shut down for at least threeweeks during the year. Virtually all technologies have installation times of three weeks or less, meaningthat no incremental shutdown time is needed.8

3.1.1.4 Waste Feed Reduction Compliance Scenario

As described above, the model plants analysis assumes that combustion facilities are limited toinstallation of pollution control equipment in responding to the new emissions standards. In fact, somefacilities may find it advantageous to reduce the quantity of pollutants fed to the combustion unit bymodifying the composition of the hazardous waste fuels they burn. When these reductions are introduced,some facilities may be able to meet the new standards with more modest pollution control measures

Feedrate-adjusted emissions are estimated by use of partitioning factors that allow translation of9

feedrate reductions into emissions reductions. These estimates were developed as part of the model plantsanalysis performed by EPA.

This reduction could be achieved either through reductions in concentrations in waste feed or simply10

through reductions in the quantity of waste burned. While reductions greater than 25 percent couldtheoretically reduce required expenditures on APCDs for compliance, practically such reductions seemunlikely. A number of fuel blenders stated that metals reductions greater than 25 percent were not feasiblegiven currently available technology. (Conference call with Chris Goebel, National Association ofChemical Recyclers, and three member companies, February 14, 1995).

New MACT standards can also be triggered when a facility remodels in such a fundamental fashion11

that it spends over one-half of the capital cost of building a new facility.

A more detailed explanation of the model plants analysis used to determine costs for new sources can12

be found in the memorandum "MACT Costs for New Sources," prepared for Frank Behan and BobHolloway, EPA, Office of Solid Waste, prepared by Wyman Clark and Bruce Springsteen (EER), March31, 1995.


because stack gas concentrations of certain HAPs are reduced. Therefore, the model plants analysis was9

extended to consider the pollution control measures that would be required if all facilities reduced thequantity of metals (including mercury) and chlorine in the waste fed to the combustion unit by 25 percent.10

As we will discuss further below, however, relatively few facilities change model plant assignment on thebasis of the 25 percent feed reduction; therefore, the impact on engineering compliance costs is limited.

3.1.1.5 MACT Compliance Costs for New Sources

While most of this report focuses on MACT standards for existing sources, the rule also proposesMACT standards for new sources entering the hazardous waste combustion market. These standards wouldapply to both newly constructed facilities (e.g., a new commercial incinerator) as well as to BIFs thatchoose to begin burning hazardous waste. 11

EPA applied the same basic model plants approach in developing engineering costs for new sourcesas was applied for existing sources. Specifically, EPA determined the set of pollution control devices thatwould be needed to meet the standards and used the cost models discussed above to estimate engineeringcosts. These estimates were developed for each category of combustion facility and for different sizeclasses. Estimation of the new MACT costs differs from existing source cost estimation in that we mustfirst assume a set of baseline pollution controls for each combustion unit that allows the unit to meet currentregulatory standards (Resource Conservation and Recovery Act (RCRA) and the BIF rule). For all kilns,this baseline control is assumed to be a fabric filter system, while the baseline control for incinerators isassumed to be a water quench cooling tower, a wet scrubber, and a venturi scrubber. The net annual costsof the new MACT standards are then calculated as the incremental pollution control costs beyond thesebaseline control costs.12

"Summary of Penta Engineering Review of EPA/OAQPS Cost Models for Cement Kiln Air Pollution13

Control Devices," prepared for Frank Behan, U.S. EPA, Office of Solid Waste (OSW), April 5, 1995.


3.1.1.6 Caveats and Limitations

The model plants analysis contains a variety of uncertainties. The most significant include thefollowing:

! The analysis is designed to estimate national compliance behavior, not complianceat specific facilities. By its nature, the model plants analysis abstracts fromindividual combustion unit decision making and assigns compliance actions in asimplified manner. Actual compliance behavior for any given unit or facility maydiffer from the pollution control measures assumed for the assigned model plant.

! Emissions data are the product of trial burns conducted at a limited subset offacilities in each combustion sector; these are the facilities included in the modelplants analysis. Even for the included facilities, emissions information for someHAPs often is not available; as mentioned, emission reduction requirements areassigned to these facilities according to the underlying statistical distribution foreach HAP (which is based on emissions of the HAP at the facilities where data areavailable).

! The OAQPS spreadsheet models serve as the primary source of capital andoperating costs for major pollution control devices. In the case of cement kilns,industry representatives have reviewed the underlying cost models and suggest thatthey contain significant inaccuracies. These comments were provided too late forintegration into the model plants analysis, but will be considered in futurerevisions. In general, the comments assert that the spreadsheet modelsunderestimate the costs associated with certain control technologies, suggestingthat the model plants analysis may understate compliance costs.13

! CEM systems for some of the HAPs are still under development, and thereforedifficult to cost out with precision. All required CEMs are expected to be readilyavailable by the time the proposed rule is promulgated.

The report issued by PSPD is dated November 1, 1994.14

The averages are as follows: 1.8 kilns per cement facility, 2.2 kilns per LWA facility, 1.1 units per15

commercial incinerator facility and 1.2 units per on-site incinerator facility (see Exhibit 2-1.).


3.1.2 Compilation of Costs

3.1.2.1 Total Costs Assuming No Market Exit

We use the results of the model plants engineering cost analysis to develop upper bound estimatesof the total costs associated with the rule. In general, we compile these costs by adding the engineering costestimates across all affected combustion units. These costs represent an upper bound because they assumethat all units will comply with the proposed rule by installing the equipment or adopting the design,operation and maintenance (DOM) measures specified in the model plants analysis. In fact, some facilitiesmay close combustion units or stop burning hazardous waste rather than comply with the proposed rule.This will reduce the overall compliance costs relative to a simple compilation of engineering costs thatassumes all affected units comply. The economic impact analysis discussed in Chapter 4 of this reportreviews our method for estimating the number of units that stop burning hazardous wastes and the revisedestimate of total costs. Exhibit 3-4 illustrates the process for calculating both the total engineering costsand the expected total costs of the proposed rule with market adjustments.

For each combustion sector, we compiled compliance costs by first assigning compliance costs toeach combustion unit based on its model plant category, and then summing the costs for each unitcharacterized in the model plants analysis. Because the model plants analysis covers only a subset of thetotal universe of units, compliance costs were scaled up from the model plants level to the level of allfacilities. To do so, we relied on a list of all operating (i.e., permitted and interim status) hazardous wastecombustion facilities furnished by EPA's Permits and State Programs Division (PSPD). However, this14

list provides only the number of operating facilities; because facilities can have multiple combustion unitsand our costs are on a per unit basis, we need to scale by the number of units. Data are not available onthe number of units at every facility, only at facilities in the model plants analysis. Therefore, for facilitiesin a given combustion category, we assumed the average number of units found at facilities characterizedin the model plants analysis. 15

Total compliance costs allowing for market exit are also illustrated on Exhibit 3-4. This refinedvalue is presented in Chapter 4, and supplements model plant data with a variety of unit specific and sector-specific data.

The degree to which the model plants analysis covers the universe of units varies by combustionsector. Exhibit 3-5 presents information on the scaling factors used for each sector. As shown, thecoverage for cement kilns and aggregate kilns is high, with almost 90 percent of the units in these twosectors included. In contrast, the sample of commercial incinerators covers half of the universe, while thesample of on-site incinerators covers about 30 percent of the universe. This degree of coverage hasimplications for the certainty of the national cost estimates as well as the economic impact analysis (i.e.,more limited sampling implies greater uncertainty in the cost and economic impact results). Furthermore,the precise number of operating units in each combustion sector is uncertain. We have included in the

These facilities were included in the analysis because they are permitted to burn waste and could16

therefore resume waste burning at any point in time.


scaling factors several facilities that the PSPD list shows as permitted but "not operating" on November1, 1994. If these facilities have permanently discontinued waste burning, we will overstate the national16

costs of the proposed rule (as well as other impacts discussed elsewhere in this report).

Exhibit 3-5

SCALING FACTORS FOR NATIONAL COST ESTIMATES

Total Number of Total NumberUnits in Model of Units in ScalePlants Analysis Universe Factor

Cement Kilns 45 50 1.1

Lightweight Aggregate Kilns 13 15 1.2

Commercial Incinerators 17 34 2.0

On-Site Incinerators 54 184 3.4

3.1.2.2 Costs Per Unit

In addition to compiling the model plant estimates onto a national level, we also used the data tocharacterize typical costs per unit in each combustion sector. The average cost per unit is simply theaverage of capital and annual costs across all units in the model plant analysis. For example, averagecement kiln capital costs are based on the capital costs for all cement kilns in the model plants analysis.

3.2 RESULTS OF MODEL PLANTS COST ANALYSIS

In this section, we review the results of the engineering cost analysis. First, we present estimatesof compliance costs nationwide by combustion sector, as well as estimates of average costs per combustionunit. We also discuss the factors that drive the cost changes between regulatory options.

3.2.1 Compliance Costs-Engineering Estimates

The proposed MACT standards will introduce cost burdens that differ greatly by regulatory optionand by combustion sector. At the simplest level, we can first consider the total sector expenditures onpollution control measures as estimated in the model plants analysis. As discussed, this provides an upperbound estimate of the social costs of the rule because it assumes that all facilities will continue wasteburning and install the requisite pollution control measures. Exhibit 3-6 presents total annual compliancecosts nationally and by sector. As shown, total costs range from $189 million per year for the MACT floor,to $561 million per year for Option 4 where a variety of above-the-floor controls are introduced.

Note that the PSPD document used to characterize the universe of combustion facilities lists several17

facilities as "not operating" as of November 1, 1994. This implies that we may overstate national costs.


Looking across combustion sectors shows that cement kilns and on-site incinerators make up themajority of the national costs under any given regulatory option. For cement kilns, this is due to a relativelylarge number of units as well as a high cost per unit (see below). For on-site incinerators, the high costsare primarily due to the large number of combustion units. Total costs are less for commercial17

incinerators (because of relatively limited costs per unit) and for LWAKs (because of the limited numberof units).

Looking across options, several patterns are noteworthy:

! Achieving the floor is between 67 and 91 percent of the total costs of compliancefor all options not requiring improved combustion. This is because achieving thefloor levels requires the bulk of the expenditures on new APCDs needed even tomeet many ATF levels.

! Adding the above the floor dioxin standard (from 0.5 to 0.2 dioxin/furan toxicequivalent (TEQ)) and improved cement kiln mercury standard (from 40 to 30ug/dscm) under Option 1b adds about 10 percent to the national cost of the rule.

! Adding the protective dioxin and mercury standards for all facilities (Option 1c)increases national compliance costs by approximately 31 percent relative to thefloor (about 19 percent relative to Option 1b). These increases are driven by theaddition of carbon systems to many combustion units.

! The improved PM control for cement kilns and more stringent acid gas controlrequirements for LWAKs under Option 3 add less than one percent to the nationalcost of the rule, because few units need to purchase incremental controls.

! The most prominent changes occur when improved combustion standards areintroduced under Options 2 and 4. For example, moving from Option 1c to Option2a increases national costs by more than 100 percent. While costs for all sectorsincrease, the major change occurs for cement kilns, where the cost of improvedcombustion far outstrips the cost of other emissions control measures (underOptions 2a and 4). These large increases are due to the use of large quantities offuel to heat afterburners.

EXHIBIT 3-6

TOTAL ANNUAL COMPLIANCE COSTS (millions)(Assuming no Market Exit)

On-siteCommercialLWACementTotalIncineratorsIncineratorsKilnsKilnsOptions

$189$94$21$9$65Original Floor

$207$103$24$9$71Option 1b

$247$117$27$12$91Option 1c

$558$128$31$32$366Option 2a

$282$128$31$32$91Option 2b

$249$117$27$14$91Option 3

$561$128$31$35$366Option 4

Notes:Compliance costs include CEM costs.1.Range of CEM costs per unit:2.

$125,509minimum$132,914maximum

Note that this table does not take into account CEMs.18


! Continuous emissions monitors play a major role in the overall costs of the rule.CEM costs add $37 million to the total annual costs of the rule and do not changebetween regulatory options.

The patterns in the national (and per-unit) compliance costs are due to changes in the pollutioncontrol measures that must be implemented to meet the proposed MACT standards. While the changesamong regulatory options are complex, Exhibits 3-7 and 3-8 present information that helps identify thecontrol technologies driving the cost changes. Exhibit 3-7 demonstrates how usage of different controlmeasures changes, showing the percentage of units in a given combustion sector that must introduceparticular technologies to meet the standards. Exhibit 3-8 presents similar data, but shows the percentageof engineering compliance costs that are attributable to each technology or control measure.18

To reach the MACT floor control requirements, about half of all cement kilns and incinerators mustadd carbon systems. These systems control emissions of mercury, dioxins, and furans. About half of theunits in all sectors must add fabric filters to control PM or metals, or to capture carbon used in carbonsystems. Other control measures are specific to each combustion sector. For example, quench systemsfor dioxin control are introduced at cement kilns, packed tower systems at incinerators, and wet scrubbersat LWAKs. On a cost basis (Exhibit 3-7), the most significant devices are carbon systems and fabric filtersfor cement kilns; afterburners, carbon systems and fabric filters for incinerators; and fabric filters andscrubbers for LWAKs.

The increase in compliance costs across Options 1b and 1c are the result of dioxin and mercurycontrol measures. For example, note that under the floor, 47 percent of cement kilns employ carboninjection; this increases to 91 percent under Option 1c (with increased mercury and dioxin standards).Likewise, expenditures on carbon injection rise from 19 to 29 percent of pollution control costs (Exhibit3-8). Because fabric filters are used in tandem with carbon injection, they are increasingly employed aswell.

The primary change under Option 3 is the introduction of more stringent acid gas controls forLWAKs. This increases the usage of spray towers to nearly 70 percent and enhances the role of spraytowers in LWAK pollution control expenditures. The increased PM control requirement for cement kilnsadds little to costs because fabric filters have already been installed as an element of mercury and dioxincontrol (to capture particles from the carbon injection system).

Compliance costs in Options 2 and 4 are driven by the cost of afterburners for improvedcombustion. For example, under Option 2a and 4, all cement kilns must install afterburners. Exhibit 3-8shows that afterburners go from being zero to 74 percent of engineering compliance costs. This samepattern applies to LWAKs. This is due to the operating costs of the afterburners and the large gas flow thatmust be handled at cement kilns and LWAKs. Operating costs for afterburners are driven by the cost offuel for firing the afterburner.


3.2.2 Average Costs Per Combustion Unit

Considering annual costs per combustion unit further highlights the variation in costs across sectorsand regulatory options. Exhibit 3-9 presents average annual costs per combustion unit. Except in oneregulatory option (Option 2b), cement kilns incur the greatest costs per unit -- typically between $1 millionand $2 million. As shown, average per-unit costs for other sectors are roughly half the per-unit costs forcement kilns under most regulatory options. Again, these costs are primarily attributable to the large sizingof the APCDs to handle the large volume of stack gas that must be controlled at cement kilns.

As with the aggregate costs, per-unit costs are greater in those regulatory options with morestringent standards. For example, the average per unit costs at a commercial incinerator increase from$617,000 to $800,000 as above the floor (ATF) levels for dioxin and mercury are introduced (i.e., movingfrom the floor to Option 1c). As shown, per-unit costs are greatest in the options (Options 2 and 4) thatrequire ATF controls for CO and hydrocarbons (indicators of complete combustion).

The variability in per-unit costs is limited by the constant flat cost of CEMs that apply to each unit.On a per-unit basis, CEMs add about $130,000 per year to the cost of compliance at each unit (under CEMoption 2).

3.2.3 Cost Reductions Associated with Feed Reduction

Exhibit 3-10 demonstrates that the 25 percent feed reduction compliance option has little effect onnational costs. For example, under the floor, national costs fall from $189 million to $180 million, adecrease of only four percent. Overall cost reductions are equally limited for other regulatory options. Theprimary reason, as will be discussed later in this report, is that relatively few combustion units are able tomeet the standards by feed reduction alone; very often full APCD trains are still needed. As a result, thechange in overall engineering compliance costs is relatively minor.

3.2.4 MACT Costs for New Sources

Exhibit 3-11 presents the incremental annual cost per combustion unit to comply with the newMACT standards (i.e., compliance costs beyond those associated with existing RCRA and BIF standards).The costs shown include pollution control equipment, plus CEMs for CO, HC, PM, and Hg as describedin Section 3.1.1.2. The estimated compliance costs for new sources are generally on par with those forexisting sources since the new MACT standards in Options 1 and 3 are similar to the more stringentstandards for existing sources. Costs for cement kilns under new MACT Option 2 are significantly higherthan those for any of the existing sources because of the strict requirements for low volatile metals (30ug/dscm).

Consistent with the existing source standards, cement kilns and LWAKs face higher compliancecosts per unit because of their large gas flow rates. For cement kilns, costs are especially high underOption 2 where afterburners must be installed to meet the stringent carbon monoxide (50 ppmv) andhydrocarbon (5 ppmv) requirements.

EXHIBIT 3-9

AVERAGE TOTAL ANNUAL COMPLIANCE COSTS PER UNIT(Assuming no Market Exit)

On-siteCommercialLWACementIncineratorsIncineratorsKilnsKilnsOptions

Estimated Number of184341550Combustion Units

$508,834$617,021$579,942$1,290,676Original Floor

$559,008$713,245$579,942$1,411,732Option 1b

$634,662$799,632$771,843$1,799,664Option 1c

$693,335$917,020$2,142,555$7,265,445Option 2a

$693,335$917,020$2,142,555$1,799,664Option 2b

$634,662$799,632$949,174$1,799,664Option 3

$695,272$917,020$2,325,167$7,265,445Option 4

Notes:Compliance costs include CEM costs.1.Range of CEM costs per unit:2.

$125,509minimum$132,914maximum

EXHIBIT 3 - 10

EFFECT OF 25 PERCENT WASTE FEED REDUCTIONON TOTAL ANNUAL NATIONAL COSTS

(millions)Percentage25 PercentNo FeedDifferenceFeed ControlControl

4%$180$189Original Floor

2%$203$2071b

5%$235$2471c

2%$547$5582a

5%$268$2822b

4%$239$2493

2%$547$5614

Notes:Compliance costs include CEM costs.1.


Large and medium sized incinerators face higher costs under new source MACT Option 3. Thisis primarily due to the cost of carbon beds that are needed to meet the stringent dioxin and mercurystandards that would apply. For small incinerators, the cost of meeting the new MACT standards does notchange between regulatory options because the same control devices are used in all cases.

3.3 CONCLUSIONS

We have employed a model plants approach to estimate the compliance costs for the proposedMACT standards for hazardous waste combustion facilities. Under this approach, individual facilities areassigned to a model plant category on the basis of current emissions and types of new air pollution controlequipment required to comply with the proposed rule. The model plant designation represents the set ofpollution control measures that will be necessary to meet the standards for each HAP in a given regulatoryoption. From this assignment of pollution control devices, we can derive the capital and operating coststhat each modeled facility would incur in complying with the proposed rule. The estimates of compliancecosts also include the costs associated with continuous emissions monitoring that may be required.

The model plants analysis yields the following conclusions:

! If all facilities continued waste burning and complied with the rule, expendituresto reach the MACT floor requirements would be approximately $189 million peryear. Complying with the above-the-floor dioxin requirements in Option 1b wouldincrease annual costs to approximately $207 million. Cement kilns and on-siteincinerators account for most of these costs.

! Per unit costs under these options are greatest for cement kilns (about $1.5 to $2million) because of the large gas flows that must be managed. Per unit costs forother facility types are about half of those for cement kilns. CEMs contributeabout $130,000 to annual per unit costs for all combustor types.

! In meeting the floor and Option 1b standards, costs would be driven by installationof carbon injection/bed systems and fabric filters at most facilities, as wellafterburners and scrubbers at some facilities.

! Total costs increase greatly when improved combustion is required (Options 2 and4). These costs are primarily driven by the need for afterburners at cement kilns.

! Only a limited number of units could reduce expenditures on pollution control byreducing waste feed. Therefore, the reduction in total engineering costs is onlyabout two to five percent.

EXHIBIT 3-11

PER-UNIT INCREMENTAL ANNUAL COSTS FOR NEW SOURCES

Regulatory Options321Category

Cement Kilns$2,336,500$5,324,500$1,755,500Small$4,986,400$12,005,400$3,445,400Large

LWAKs$1,495,900$1,434,900$1,434,900Medium

Incinerators$336,000$336,000$336,000Small$720,500$514,500$514,500Medium

$1,188,700$772,700$772,700Large

Note:CEM costs assume CEMs for CO, HC, PM, and Hg. CEM costs at new facilities are assumed to be equalto the costs at existing units.

Sources:"MACT Costs for New Sources," memorandum to Frank Behan, EPA/OSW, from Wyman Clark and BruceSpringsteen, EER, March 31, 1995."MACT Compliance Costs," memorandum prepared for Larry Denyer, EPA, prepared by Greg Kryder,Wyman Clark, EER, April 8, 1995; revised April 24, 1995.


! Per-unit compliance costs for new sources are on par with the costs for existingsources. Under all regulatory options, cement kilns would incur the greatest costs;this is especially true in new source Option 2 where more stringent PIC controlsare introduced.

The engineering costs discussed here provide only an upper bound estimate of the total costs of theproposed rule because they assume that all facilities will comply with the regulation and continue wasteburning. The unit-specific cost estimates from the model plants analysis serve as a primary input into oureconomic impact analysis in Chapter 4 where we arrive at a revised estimate of total costs.

4-1 DRAFT: November 13, 1995

ECONOMIC IMPACT ANALYSIS CHAPTER 4____________________________________________________________________________________

While Chapter 3 presents an upper bound estimate of the total economic costs of the proposedhazardous waste combustion maximum achievable control technology (MACT) standards, actual costs willdepend upon the incentives and reactions of the regulated community. In this chapter the EnvironmentalProtection Agency (EPA) draws on a diverse set of data to characterize the economics of hazardous wastecombustion and estimate how increased compliance costs would affect the incentive that combustionfacilities have to continue burning and the competitive balance in combustion market segments. Inaddition, the Agency considers the costs that would be imposed on groups other than combustion facilities-- particularly, waste generators and fuel blenders. The overall objective is to develop a refined estimateof the total costs that will actually be incurred by the regulated community.

The discussion below is divided into methodology and results. The methodology section introducesa screening process used to assess economic impacts. In addition, we describe how these screens can becombined with information on alternatives to combustion (i.e., waste management alternatives and wasteminimization) and data from the model plants analysis to assess how waste flows might change when newcombustion standards are introduced that increase the cost of combustion. The results section presents andinterprets these screening measures and discusses how the reactions of the regulated community wouldaffect the total costs of the rule and the flow of waste.

Overall, the economic impact analysis indicates that some combustion facilities may experiencea substantial change in the cost of burning waste, but that this change is likely to have a limited impact oncombustion markets. In terms of effects on waste-burning cost structure, cement kilns and lightweightaggregate kilns (LWAKs) are most affected by the regulation. This is primarily a product of their relativelylow baseline costs of burning, meaning that incremental compliance costs represent a large increase in theiroverall cost of burning waste. For incinerators, compliance costs are lower, represent smaller additionsto baseline costs, and change little across regulatory options. The analysis concludes that cement kilns havethe lowest waste burning costs even after regulation, and so will be able to set the market price.

The availability of waste management alternatives other than combustion may constrain themagnitude of the price increase passed to generators. Because there are limited economic data on wasteminimization and other non-combustion waste management alternatives, EPA was unable to predictprecisely how much combustors could raise prices in the face of higher compliance costs. The Agencytherefore conducted a sensitivity analysis, using a low and a high price pass-through scenario, as well asbounding the assessment using zero percent pass-through. The low price pass-through scenario assumes

Since price pass-through has been calculated on a pre-tax basis, and new investments can be deducted1

from taxable income, reducing their cost to the corporation, combustors would have an impetus to recoverless than the pre-tax costs per ton through price increases.


that waste minimization and non-combustion alternatives become viable at combustion prices only two to13 percent higher than the current market. The high price pass-through scenario assumes that the most1

efficient combustors could raise prices to recover much of their new compliance costs without generatorsshifting to alternatives. Zero-percent pass-through assumes that all costs of the rule are borne by existingprofits.

To the extent that compliance costs cannot be passed through to generators and fuel blenders, theprofitability of waste burning in kilns will fall. EPA estimates that if prices can be increased by less than13 percent, a significant number of facilities may cease to burn waste. However, because the facilities thatare likely to cease waste burning are those that have little waste over which to spread costs, the waste shiftsthat result are quite minor (less than five percent of all combusted waste), leaving the competitive balancerelatively unaffected. These conclusions vary little across regulatory options, except those where productsof incomplete combustion (PICs) must be controlled; under these options, kilns incur much larger costs andmay be more adversely affected.

4.1 METHODOLOGY

The methodology for assessing economic impacts is divided into several components:

! Assessment of waste management alternatives;

! Assessment of waste minimization;

! Basic economic impact indicators;

! Breakeven quantity analysis;

! Effect of waste feed reductions; and

! Effect of removing the small quantity burner exemption.

Each of these components is discussed in the sections below. Before doing so, the basic data used in thescreens and in other aspects of the economic impact analysis are discussed.

Prior to presenting the various screening analyses we use for determining economic impacts, twoissues should be noted. First, the screening analyses all seek to characterize the incentive that combustionfacilities have to burn waste and determine if combustion units will continue to operate in the face ofincreased costs. Implicit in this analysis is the assumption that combustion units operate as separate profitcenters. In fact, a chemical plant employing an on-site incinerator could subsidize the incineratoroperations with profits from some other component of the facility. Similarly, a large waste managementcompany may subsidize commercial incinerator operations with income from another line of business.


Such actions, however, would violate basic microeconomic assumptions of profit maximization. Therefore,the analysis assumes that hazardous waste combustion activities must be profitable to justify continuedoperation.

A second issue to note concerns our rationale for choosing the model plants and unit profitabilityapproach over the econometric approach sometimes used to assess regulatory impacts. The econometricapproach uses historical price and output information to predict changes in industry output given marginalchanges production costs. Two factors make this approach less desirable for assessing the economic impactof the proposed MACT standards. First, the cost increases implied for some sectors are substantial, andgo beyond the marginal cost changes most readily analyzed by the econometric approach. Second,hazardous waste combustion markets have changed rapidly over the last several years, with new entrants(e.g., cement kilns) as well as new regulations. These exogenous changes make it difficult to predictindustry behavior on the basis of historical price and output data.

4.1.1 Data Inputs for Economic Impact Analysis

Our economic impact screens are developed using facility-specific data in conjunction with anumber of industry standard values for pricing. These are combined with model plant estimates for thebaseline costs of combustion and new compliance costs at each category of combustor. As the newcompliance costs were described in detail in Chapter 3, we describe only the other key data inputs inChapter 4. Exhibit 4-1 illustrates how all of the pieces were combined to evaluate the economic impactof the proposed rule.

4.1.1.1 Facility-Specific Data and Other Key Inputs Used in Economic Screens

Evaluating the impact of the proposed standards on existing combustion units required a varietyof data inputs. Facility revenue estimates required data on waste quantities burned, energy savings fromburning hazardous wastes, and the market prices of combustion. For on-site incinerators, transportationcosts were also needed in order to estimate the avoided cost of off-site treatment. This section summarizesour key data sources.

Data on the total quantity of waste burned were available at the facility level for all combustionsectors. The 1991 Biennial Report Survey (BRS) provided quantity data by source and form code for eachwaste stream burned at an on-site incinerator. To match the waste streams with the available pricing data,we grouped them by general type of waste (i.e., liquids, sludges and solids). The waste streams withunknown source and form codes were distributed among the other waste types, assuming the samepercentage of liquids and solids for each facility.

Data for the commercial incinerator sector are from Jon Hanke, "Hazardous Waste Incineration 1994,"2

EI Digest, June 1994, p. 25. Data for the cement and lightweight aggregate kiln sectors are from JeffreySmith, "Industrial Furnaces 1994," EI Digest, October 1994, p. 20.

For the commercial incinerator sector, we used Office of Solid Waste, U.S. Environmental Protection3

Agency, Background Document for Capacity Analysis for Newly Listed and Hazardous Debris to Support40 Code of Federal Regulations (CFR) 268 Land Disposal Restrictions (Final Rule), Volume 1: CapacityAnalysis Methodology and Results, June 1992, Exhibit 2-5. Data for the cement and lightweight aggregatekiln sectors are from Office of Solid Waste, U.S. Environmental Protection Agency, Background Documentfor Capacity Analysis for Land Disposal Restrictions Phase II - Universal Treatment Standards, andTreatment Standards for Organic Toxicity Characteristic Wastes and Other Newly Listed Wastes (FinalRule) Volume I: Capacity Analysis Methodology and Results , August 1994, Exhibit 2-7.

We inflated the 1993 prices to 1994 prices using the Gross Domestic Product (GDP) implicit price4

deflator for services reported in the Survey of Current Business, Table 7.1. Pricing data are from Hanke,op. cit., p. 23.

Jeffrey Smith, "Fuel Blenders 1994," EI Digest, September 1994, p. 28.5


More recent data were available for the commercial sector from EI Digest, a trade journal. EI2

Digest surveyed the commercial incinerator, cement kiln, and lightweight aggregate kiln sectors forinformation on the quantity of waste burned. As these data represent total quantities and are not brokendown by general type of waste, we extrapolated from other data sources to estimate the percentage ofliquids, sludges, and solids burned at each facility. For facilities for which there were no available data,we assumed the average mixture of liquids, sludges, and solids for each sector. The available data on themixture of waste types burned at each facility were more recent for the cement and lightweight aggregatekiln sectors than for the commercial incinerator sectors. 3

Facilities with more than one unit permitted to burn hazardous waste often reported waste quantitiesburned at the facility-level rather than by combustion unit. In such circumstances, we evenly split the totalquantity of waste burned among all the permitted units. Although this method does not take into accountthe capacity of each unit, data on the practical waste burning capacity were not available at the combustionunit-level.

The revenues commercial combustion units earned from burning hazardous waste were estimatedby multiplying the quantity of waste combusted by the average prices that facilities charge to combust thewaste. As pricing data on sludge wastes were not available, we assumed that commercial incineratorscharged generators the same price for sludges as they did for liquids ($293/ton). Because liquids prices4

are significantly lower than solids prices, this represents a conservative assumption that could overstate theimpact of the rule on commercial incineration facilities. The prices used to calculate the revenues earnedby cement and lightweight aggregate kilns were the average price that fuel blenders paid to kilns. Theseprices ranged from $100/ton for liquids to $740/ton for solids.5

DPRA, Incorporated, September 1994. Data were inflated to 1994 prices using the GDP implicit price6

deflator.

We used the average Btu/lb estimates used in the baseline cost models. These models assumed 13,1117

Btu/lb of liquids burned by cement kilns and 10,767 Btu/lb of liquids burned by lightweight aggregate kilns.For sludges and solids burned by both types of kilns, we used an average heat content of 9,733 Btu/lb.

U.S. Cement Industry Fact Sheet (12th edition), Table 24.8

We assumed the price of coal for Cost, Insurance, and Freight (CIF) Electric Utility Power Plants (Table9

7.8). For natural gas, we used the price for industrial customers (Table 6.9). For residual fuel oil, the priceto end users (Table 5.21). We inflated preliminary 1993 prices to 1994 prices using the GDP implicit pricedeflator for services.


The costs on-site incinerators avoid by not shipping the hazardous waste off-site were estimatedby multiplying the quantity of waste by the price they would pay to ship the waste to a commercialincinerator. Both the fee charged by the commercial incinerator and the cost of transporting the waste wereincluded. Assuming an average distance of 200 miles, the cost of transporting liquid waste to a commercialincinerator was estimated to be $51/ton. The cost of transporting sludges and solids was estimated to be$49/ton.6

In addition to the revenues facilities earn from combustion fees, we estimated the savings to cementand lightweight aggregate kilns from avoided energy purchases. These are the result of recovering theheating value of the hazardous waste fuels burned in their kilns. To calculate energy savings, we firstconverted the waste quantities burned into an energy equivalent (in million Btus per pound). The energy7

content of the waste fuels was then compared to the energy content of conventional fuels to calculate theconventional fuels displaced by waste burning.

The conventional fuels burned in the kilns varied by sector. For cement kilns, we used the averagemixture of coal and natural gas burned in the industry, as reported by the Portland Cement Association.8

For lightweight aggregate kilns, we assumed the kilns would replace the hazardous waste they burn withresidual fuel oil. Data on the energy content and prices of these conventional fuels were taken from the9

1993 Annual Energy Survey.

These data provide key inputs to our combustion model. While the model provides importantinformation on differential impacts of the proposed regulation on particular combustion sectors, it is notintended to precisely predict the impact of the rule on specific combustion units. The results are sensitiveto the quality of several key inputs: quantity of wastes burned per unit; the mixture of solids and liquidsburned at each unit; and pricing of combustion services. Data in all cases are several years old. Pricingdata use market averages, and therefore do not reflect variability due to geography or to the specific wastestream being burned. Were more recent data substituted into the model, some facilities that appear non-viable may turn out to be healthy, and vice-versa. However, such improvements to the approach are notpossible using current publicly-available data.


4.1.1.2 Baseline Cost Analysis

To evaluate the impacts of the proposed rule on hazardous waste combustors, baseline costs ofcombustion in the current marketplace must first be established. Baseline costs suggest importantdifferences across combustion segments that significantly influence competitiveness. The results of thebaseline cost analysis provide a core input to the combustion cost model. Below, we summarize how thesebaseline costs are estimated. A more detailed description of the approach, as well as detailed results, canbe found in Appendix B.

The objective of the baseline cost analysis is to estimate the total costs (variable and fixed) ofburning a ton of hazardous waste in combustion units of different types. In the case of incinerators, thisbaseline cost is simply the variable and fixed costs of the facility (prior to new pollution controlrequirements), since incineration is the sole function of the facility. For cement kilns and commercialboilers and industrial furnaces (BIFs), the decision is whether to burn hazardous waste or some other fuel.In this case, we need to know the incremental costs introduced by the decision to burn hazardous wasterather than conventional fuel; this is the cost that would be avoided if the facility chose to burn conventionalfuel. These incremental costs might include permitting costs, the cost of insurance, and the cost of specialhazardous waste handling procedures and equipment. Because the same kiln is required for cementproduction regardless of hazardous waste combustion activities, no kiln capital costs are included in thebaseline cost estimates for cement kilns.

The baseline cost analysis involved four key tasks:

! Identification and classification of combustion cost components;

! Development of model plant categories;

! Quantification of combustion cost components; and

! Development of annualized baseline combustion cost estimates for each modelplant category.

EPA first identified the key elements of baseline costs for kilns and incinerators. For cement kilns,key cost components include waste storage, waste sampling and analysis, and waste-specific labor. Forincinerators, key components include the cost of the combustion unit and air pollution control device(APCD) units already installed, labor, and incinerator ash disposal. Annualized permitting costs arerelatively more significant for BIFs than for incinerators.

We then classified the baseline cost components into three categories: fixed annual capital; fixedoperating and maintenance costs (O&M); and variable costs. Fixed annual capital costs refer toexpenditures lasting multiple years. This includes capital equipment and operating permits. Costs havebeen annualized using a 10 percent interest rate to convert the total capital cost to a series of equal annual

CRF ' i (1 % i)(1 % i)n & 1

, where i ' 10% and n ' 10, 15, or 20 years.

A 10 percent real rate of return was used to calculate a capital recovery factor (CRF) using the following10

equation:

This assumption leads to lower annual O&M costs, reducing the cost per ton combusted.11

The distinction between fixed O&M and fixed capital is important in our calculation of short-run12

breakeven quantities. While fixed capital is sunk and need not be recovered for a unit to continue burningwaste, fixed O&M is a recurring cost and must be recovered through revenues.


payments over the estimated life of the capital. Fixed O&M costs include items such as annual machine10

repairs. These costs recur every year, but do not vary significantly in proportion to the quantity ofhazardous waste burned. Variable costs include items such as some labor costs that increase in proportionto the amount of waste burned. Annual variable costs are derived by multiplying variable costs per ton ofwaste burned by the number of tons burned.

After identifying the key cost components to include in the baseline analysis, model plantclassifications were developed to characterize the current combustion universe. Model plants weredeveloped for each industry sector, including commercial incinerators, cement kilns, lightweight aggregatekilns, and on-site incinerators. Within each sector, additional model plants were developed where costdifferences across combustion units were large. Unit type (e.g., wet kiln versus dry kiln); unit size; andinstalled APCD train all affected model plant classifications. Six model plants were developed forcommercial incinerators; 12 for on-site incinerators; four for cement kilns; and three for lightweightaggregate kilns. The model plants in the baseline cost analysis are based on the type and size of thecombustion unit. They differ from the model plants developed to estimate the new cost of compliance thatare described in Chapter 3 of this report.

As with any modeling, the model plants used in the baseline cost analysis are approximations ofactual units rather than precise replications. Model plants are assumed to operate continuously for everysector except on-site incinerators. On-site incinerators are assumed to operate in batch mode because theyare generally small, combust relatively small quantities of hazardous waste, and would consume a greatdeal of energy if they were to be operated continuously. On-site incinerators are also assumed to burn11

only hazardous wastes. To the extent that non-hazardous wastes are also burned, the fixed costs per tonof hazardous waste burned would decline.

A number of sources were used to quantify baseline cost components. These included tradepublications, engineering cost models, and best engineering judgment. The sources for each componentare detailed in Appendix B. The components were then compiled into a cost model that divided the costsinto fixed and variable costs per ton of hazardous waste combusted. We have separated annual capitalrecovery figures shown in Appendix B from the other annual fixed costs in the exhibit below. This isbecause annual fixed O&M costs would cease if a unit stopped combusting hazardous waste, while capitalcosts apply to equipment already purchased and therefore could not be recovered. "Adjusted" variable12

costs are shown for the on-site incinerator sector and reflect the batch operations of the incineratorsmentioned above.


Baseline combustion costs by model plant category are summarized below in Exhibit 4-2. Withthe exception of some on-site liquid injection incinerators, baseline costs of hazardous waste burning aresubstantially lower for BIFs than for incinerators. In all sectors, larger units have a lower fixed cost perton of capacity. These economies of scale illustrate the importance of capacity utilization; a large facilitycan have extremely high costs per ton of waste actually burned if much of its combustion capacity is notbeing utilized.

4.1.1.3 Baseline Operating Profits

To provide additional perspective on the baseline competition between different combustionsectors, EPA calculated the average baseline operating profits per ton of waste in each sector. Operatingprofits are calculated as follows:

Operating Profits = Waste Burning Revenues - Waste Burning Costs

Where:Waste Burning Revenues = Weighted average price per ton for combustion services + Weightedaverage avoided energy costs per ton of waste burned + Any price increases to recover newcompliance costs (zero in the baseline)

Waste Burning Costs = Average baseline costs per ton to burn hazardous waste + Average newcompliance costs per ton (zero in the baseline).

Operating profits are calculated before tax and deductions for plant and corporate overhead. After-taxprofits would be lower.

Baseline operating profits by sector are shown in Exhibit 4-3. The combustion units in all sectorsthat burn very little waste have negative baseline operating profits before consolidation. Cement units havelower operating profits per ton on an absolute dollar basis than commercial incinerators, reflecting the factthat they burn lower-priced liquid wastes. However, the kilns have higher average percentage operatingprofit margins due to their low baseline cost of burning. Both LWAKs and on-site incinerators have manyunprofitable units as well as many profitable ones.

4.1.2 Method for Analyzing Waste Management Alternatives

Under the proposed regulation, combustion costs and prices will rise. As the price of combustionrises, new alternatives to combustion become competitive. The greater the tons of waste currentlycombusted that can be managed less expensively in some other way, the larger the impact of wastediversions is likely to be on combustion facilities, and the more difficulty combustors will have raisingprices.

EXHIBIT 4 - 2

BASELINE COST PER TON OF PRACTICAL CAPACITY OF HAZARDOUS WASTE BURNING(Excludes Energy Savings for BIFs)

Implied TotalCost/ Ton of Fixed Costs/PracticalHours of PracticalAnnualized O & MAnnualized

PracticalTon of PracticalCapacityOperation/CapacityVariable/ TotalTotalCapacityCapacity(tons/year)Year(lbs/hour)Ton of WasteFixedCapitalAssumed APCDKiln TypeGroup

[7+3] (Note)7=[1+2]/66=[4*5]/200054321

54.7126.7147,2008,00011,80028617,445643,211ESPWet kiln (S)1Cement kilns45.0414.0494,0008,00023,50031622,319697,628ESPWet kiln (L)271.9029.9036,8008,0009,20042579,291521,094ESPDry kiln (S)346.6420.6462,4008,00015,60026614,513673,612FFDry kiln (L)4

98.6762.6712,0008,0003,00036397,321354,757FFLWAK (M)1LWAKs87.5150.5116,0008,0004,00037410,402397,751FFLWAK (M)262.4835.4826,0008,0006,50027438,991483,501FF/VSLWAK (M)3

227.3290.3251,7088,00012,9271371,662,5133,007,622FF/PBSRot. kiln (M)1Commercial265.85100.8551,7088,00012,9271651,795,4873,419,121SD/FF/PBS/IWSRot. kiln (M)2Incinerators331.8981.8982,7168,00020,6792502,194,1384,579,105FF/STRot. kiln (L)3382.0490.0482,7168,00020,6792922,357,5315,089,841SD/Q/PBS/ESPRot. kiln (L)4207.1240.1251,7208,00012,9301671,002,9861,072,120Q/FF/VS/PBSLiq. Inj. (M)5176.4938.4951,7208,00012,930138980,5601,010,277Q/PBSLiq. Inj. (M)6

192.5586.554,9408,0001,235106102,191325,348Q/PBSLiq. Inj. (S)1On-site220.88102.884,9408,0001,235118119,955388,268Q/VS/PBS/ESPLiq. Inj. (S)2Incinerators69.5925.5936,7928,0009,19844245,105696,280Q/VS/PBSLiq. Inj. (M)362.6224.6236,7928,0009,19838237,269668,534Q/PBSLiq. Inj. (M)446.7014.70102,2528,00025,56332416,5301,087,055Q/PBSLiq. Inj. (L)542.3714.37102,2528,00025,56328408,6421,060,439Q/VSLiq. Inj. (L)6788.01524.012,4088,000602264313,271948,536WHB/FFRot. kiln (S)7863.89549.892,4088,000602314327,777996,355Q/WHB/VS/PBSRot. kiln (S)8223.49118.4932,3648,0008,091105975,0742,859,771Q/VS/PBS/IWSRot. Kiln (M)9206.11118.1132,3648,0008,09188972,3472,850,113Q/PBS/ESPRot. kiln (M)10156.0985.0962,3048,00015,576711,382,6353,918,801VSRot. kiln (L)11167.3990.3962,3048,00015,576771,458,4684,173,246Q/IWSRot. kiln (L)12

Notes:On-site incinerator costs uses costs adjusted for batch feeds.1.APCD Abbreviations are as follow: ESP = electrostatic precipitator; FF = fabric filter; IWS = ionizing wet scrubber; PBS = packed bed scrubber; Q = quench; SD = spray dryer; ST =2.spray tower; WHS = waste heat boiler.Practical capacity is a measure of the maximum waste burning capacity available. Because most combustion units burn lower quantities in actuality, baseline costs per ton of waste3.burned will be higher (see Exhibit 4-6).

Matthew Gardner, et al., Energy and Environmental Research Corporation, "Development of Baseline Costs for Hazardous Waste Incineration," prepared for U.S. EPA, April 18, 1995.Sources:

Exhibit 4-3

BASELINE OPERATING PROFITS PER TON OFHAZARDOUS WASTE BURNED

(Number of Combustion Units Falling in Range)

>$150$101 - $150$51 - $100$0 - $50<$0

12315156Cement Kilns

120012LWA Kilns

1604014Commercial Incinerators

587171092On-site Incinerators

BASELINE OPERATING PROFITS AS A PERCENTAGE OF BASELINE WEIGHTED AVERAGE PRICES PER TON

(Number of Combustion Units Falling in Range)

>50%26% - 50%11% - 25%0% - 10%<0%

3010406Cement Kilns

400012LWA Kilns

866014Commercial Incinerators

5514141092On-site Incinerators

Notes:Baseline Operating Profits = (weighted average price per ton + weighted average1.energy savings per ton) - total annual baseline costs per ton. Total annual baselinecosts include fixed annual capital costs, fixed annual operating and maintenancecosts, and annual variable costs.Baseline operating profits exclude overhead, other administrative costs, and taxes. 2.Actual after-tax profits will be lower.Number of units with average operating profits less than $0 (or <0%) includes3.those burning very little or no waste.Baseline operating profits are calculated at the unit level. Consolidating burning into4.fewer units may reduce facility closures, explaining why the unit estimates presentedin this exhibit appear higher than the facility closures presented in later exhibits.Includes combustion units not currently burning waste in the cement kiln, LWAK,5.and commercial incinerator sectors; or burning less than 50 tons per year in theon-site incinerator sector.

Jerome Strauss, "Memorandum: Evaluation of Alternative Treatment Technologies for Wastes that Are13

Currently Being Sent to Combustion Facilities (Revised-Draft)," prepared for Industrial Economics, Inc.,March 30, 1995.

Remediation and other "one-time" wastes comprised about 15 percent of total combusted wastes in 1991.14

See U.S. EPA, Setting Priorities for Waste Minimization, July 1994, p. 2-12.


Our evaluation of waste alternatives involved five steps:13

! Characterizing data on waste shipments to combustion.

! Identifying the key waste streams relying on combustion.

! Identifying alternative treatment technologies for these key waste streams.

! Identifying cost information on these alternatives.

! Assessing the impact of viable alternatives might have on particular combustionsectors.

A complete description of the waste alternatives analysis is included as Appendix D.

Data characterization involved examining the Biennial Report Survey (BRS) data base to groupcombusted wastes by source and by waste form (e.g., liquid, solid). In this manner, data reported bygeneration unit were grouped by waste type so that total quantities burned for each type could be calculated.BRS entries containing insufficient source or form information to allow classification were omitted. Inaddition, gases were also excluded because BRS contained insufficient descriptions to identify alternativetreatment options. Finally, remediation and closure wastes were omitted because they represent one-timewaste sources rather than a continuing source of waste generation.14

From the remaining source code/form code groupings, the ten largest quantity groupings werechosen as the key waste streams. The choice was made on the basis of tonnage because the diversion oflarge waste streams from combustion will have the greatest impact on the viability of the combustors.These ten source code/form code groupings comprised nearly 80 percent of the total wastes incinerated in1991. The Vendor Information System for Innovative Treatment Technologies (VISITT) and theAlternative Technology Information Center (ATTIC) databases, as well as recent work done by EPA, wereused to identify waste management alternatives for these key waste streams.

Once the alternative treatment technologies were identified, treatment-specific information (i.e.,waste-feed characteristics, treatment levels, and cost factors) was collected and summarized. Engineeringjudgment was used to evaluate the applicability of particular management approaches to the key wastestreams. Limitations to the applicability of particular alternatives are also noted.

The remainder were non-routinely generated wastes and smaller quantity waste groupings. Non-routine15

wastes are unattractive options for waste minimization because the available period to recoup investmentsis short. Larger waste groupings were chosen as a focus to identify the waste minimization options likelyto have the largest impact on the market.

Payback analyses do not discount future cash flow back to present dollars, and therefore overstate the16

benefits of particular investments. Given the large uncertainties associated with the cost of wasteminimization options, the payback approach provides a reasonable first approximation of the economicdesirability of these approaches.


While the goal was to develop relatively precise cost estimates for viable alternatives, cost datawere scarce. In addition, while a number of alternatives appear to be substantially less expensive than thecurrent cost of combustion, generators continue to burn wastes. This suggests that other factors are atwork. These factors, such as liability concerns, are presented below in section 4.2.2 and incorporated intoour assessment of shifts from combustion to waste alternatives under the proposed rule. While thealternatives assessment is useful as an indicator of long-term trends, cost data are too imprecise to estimateshort-term diversions to alternatives in response to increased combustion prices.

4.1.3 Method for Analyzing the Impact of Waste Minimization

A complete evaluation of how waste management markets will react to the proposed MACTstandards must take into consideration the options available to waste generators facing increasedcombustion costs. One likely response is for generators to consider waste minimization measures thatreduce the quantity of waste needing treatment. As with alternative waste management approaches,characterizing the availability of waste minimization options will allow us to assess the elasticity of demandfor combustion services. That is, if lower cost waste minimization options are readily available for largeportions of combusted waste, combustion facilities will be less able to pass compliance costs along togenerators in the form of increased combustion prices.

EPA has developed a screening analysis to characterize the waste minimization potential for wastesthat are currently combusted. The methodology for this analysis is summarized here and explained morefully in Appendix C. EPA first compiled BRS data on combusted wastes, segmenting the data by BRSsource code (the process generating the waste) as well as SIC. Applying engineering judgment, the Agencyidentified waste minimization measures that would potentially be applicable for the waste source code andindustry in question. Roughly 1.8 million tons of waste (approximately half of all combusted waste) wereevaluated. Assignment of waste minimization technologies is subject to a significant degree of15

uncertainty given the lack of knowledge on the form of the waste, the specific constituents in the waste,and other factors affecting the applicability of specific waste reduction technologies.

To evaluate whether facilities would adopt applicable waste minimization measures, a simplifiedpayback analysis was used. Using information on per-facility capital costs for each technology, EPA16

estimated the period of time required for the cost of the waste minimization measure to be returned inreduced combustion expenditures. For example, if a waste minimization measure costs $100,000 to installand eliminates 100 tons of waste per year; and if the cost of combustion is $500 per ton, the payback periodis: $100,000/($500*100) = 2 years. Several key assumptions were applied in the payback analysis:

Prices are based on EI Digest, June 1994, p. 23. Because the specific form of the waste is unknown and17

because the wastes are generated by a diverse set of processes and industries, it is difficult to specify thetype of combustion used to manage the waste. We have assumed commercial incineration prices areapplicable. These prices are higher than those charged by cement kilns, meaning that we may overstatecombustion cost savings, understate payback periods, and therefore overstate waste minimization potential.


! Waste minimization measures with a payback of less than three years will beimplemented. Measures with a payback period of between three and ten years willbe implemented by 50 percent of the generators. Measures with payback periodsof greater than 10 years will not be implemented. These simplified cutoffassumptions are based on best engineering judgment.

! The cost of combustion is based on an assessment of the liquid/solid ratio of thewastestream; this ratio is used to develop a weighted price of combustion for thespecific waste. Liquid and solid combustion prices are based on commercialincineration prices. The liquid price of combustion is assumed to be $284 per tonwhile the solids price is $1,335 per ton.17

! Recovered process water is re-used at the facility (thereby avoiding potentialtreatment and disposal costs).

The assessment of waste minimization yields estimates of the tonnage of combusted waste that might beeliminated. It is important to note that comprehensive data to evaluate waste minimization were notavailable. Improved information on the capital investment and operating costs associated with the wasteminimization measure, as well as additional savings to the generator from avoided raw material, labor, orenergy costs are needed, and EPA invites industry to provide such data.

Incorporating both the capital and operating costs of waste minimization into a more formaldiscounted cash flow analysis would reduce the applicability of some waste minimization options thatappear viable using the payback approach. Incorporating additional savings to the plant through avoidedraw material, energy, and waste handling costs would have the opposite effect.

4.1.4 Economic Impact Screens

The purpose of our economic impact screening measures is to provide information on the relativemagnitude of compliance costs under various regulatory scenarios and to help gauge the differential impactof the proposed MACT standards across combustion sectors. Each of the screens is designed to draw onthe data discussed above to place compliance costs in a context that allows us to determine if the costs aresubstantial enough to affect waste burning operations at the combustion facility. We rely on three basiceconomic impact screens:

Annual New Compliance CostWaste Burning Revenue

' Annual New Compliance Cost/TonAverage Waste Burning Revenue/Ton%Energy Savings/Ton

Units where available data suggest that zero tons are burned are eliminated from the calculation.18


! New compliance costs per ton of waste burned;

! New compliance costs as a percentage of waste burning revenues; and

! New compliance costs as a percentage of baseline combustion costs.

Each of these is explained below. Additional information can be found in Section 4.2.1.

4.1.4.1 New Compliance Cost Per Ton of Waste

One basic indicator that can be used to measure the economic impact of the proposed MACTstandards is the compliance cost per ton of waste burned. This indicator provides information on how muchprices per ton would have to increase if combustors are to maintain their existing profits. Obviously, thehigher the cost per ton, the more likely that alternatives to combustion, such as waste minimization, willbecome economic.

For each regulatory option, we estimate the average cost per ton by combustion sector in two steps.First, we calculate the cost per ton for each combustion unit by dividing total annual compliance costs bythe tons burned at the unit. We then estimate the average cost per ton for the sector by taking the simpleaverage across all units. As with the other screens, the cost per ton screen relies directly on tonnage data18

for each combustion unit. Uncertainty in this tonnage affects the accuracy of the cost per ton estimate. Inparticular, we use BRS data as the source for tonnage burned at on-site incinerators. This tonnage is verylow for some units, leading to extremely high costs per ton and skewing the overall sector average to a highcost per ton. This finding may not be legitimate if BRS data do not fully capture the extent of wasteburning at these units or if these incinerators also burn non-hazardous waste and are able to spread theirfixed costs over a larger quantity than is apparent from BRS.

4.1.4.2 New Compliance Costs as a Percentage of Waste Burning Revenues

The second screening metric involves examining costs as a percentage of waste burning revenues,illustrating what portion of revenues would be absorbed by the costs of the new rule in the absence of anyprice increase. For each facility in the model plants analysis, we perform the following calculation:

"Cost Structure" refers to the overall magnitude and the components of costs that a combustor incurs.19

Examples include fixed versus variable costs.


As shown, annual compliance costs per ton (based on model plant costs and tonnage burned at the unit) aredivided by revenues. Revenue is calculated by multiplying the price received per ton of waste by thetonnage burned at the unit. Using available data on average prices, we apply per-ton prices for liquid,sludge and solid waste burned at each combustion unit. Revenues per ton consist not only of the price theunit receives per ton of waste, but also any savings in conventional fuel that the unit realizes (for cementkilns and LWAKs). These savings are created because the energy value of burning waste allows reducedexpenditures on fuel such as coal.

If compliance costs are large in comparison to waste burning revenues, it is more likely that thenew costs will affect waste burning behavior. Because of its simplicity, this screen does not definitivelycharacterize a combustion unit's economic incentive to continue waste burning. However, it does providecontext for determining if the costs of waste burning exceed the direct financial benefits to the combustor.

Note that this indicator relies on the concept of "imputed revenues" at on-site incinerators. Thesenon-commercial burners do not actually receive revenue when they burn waste on-site. What on-siteburning saves, however, is the cost of sending waste off-site to a commercial combustion facility. Thiswould include both the commercial combustion price as well as the avoided cost of transporting waste toan off-site facility. These savings can be treated as implicit revenue for the purposes of comparingcompliance costs to waste burning revenues.

This screen has two key uncertainties that should be taken into consideration:

! Data on current quantities burned are somewhat limited, especially for on-siteincinerators. The tonnage burned information directly affects the calculation ofrevenues.

! Prices are based on national averages. Geographic differences in pricing andtransport distances reduce the accuracy of the screen and affect the revenueestimates.

4.1.4.3 New Compliance Costs as a Percentage of Baseline Costs

Comparing new compliance costs to the baseline costs of waste burning at different facility typesprovides a measure of how much a combustor's cost structure will change from the rule. The larger the19

change, the more prices charged will rise and/or profits on waste burning will fall. For each combustionunit, we compared the model plant's compliance costs to the assigned baseline cost of waste burningdeveloped in the baseline cost analysis.

The estimation of costs as a percentage of baseline costs is subject to both the uncertainties of themodel plants analysis (see Chapter 3) as well as the uncertainties surrounding the estimation of baselinecosts, as discussed above.

For additional information on breakeven analysis, see Eugene Brigham and Louis Gapenski, Financial20

Management Theory and Practice, 6th Edition, (Chicago: The Dryden Press), 1991, p. 483; or LeopoldBernstein, Financial Statement Analysis: Theory, Application and Interpretation, (Howewood, IL: Irwin,1983), pp. 640-652.

As noted in Section 4.1, some firms could decide to operate their combustion unit at a loss. We21

anticipate that the vast majority of combustion units will shut loss-making operations, however.


4.1.5 Breakeven Quantity (BEQ) Analysis

The BEQ measures the quantity of waste that a combustion unit would have to burn to cover thecosts of operation. We use estimates of breakeven quantity to assess the likelihood that combustion20

facilities will stop burning waste in the face of increased compliance costs. We calculate two BEQmeasures -- short run and long run. Combustion units will continue to operate in the short-run if they canburn enough waste to cover their variable and fixed O&M costs. Units must cover their fixed capital costsas well if they are to continue operating in the long run. In both the long and short run, a combustor willnot choose to invest in new capital (i.e., pollution control equipment) unless it is confident that it can burnenough waste to cover the cost of that new equipment. 21

The methodology for the basic breakeven quantity measure is explained below, followed by adescription of how the BEQ is refined to arrive at an improved estimate of likely market exit.

4.1.5.1 Breakeven Quantity Calculation

In its simplest form, the breakeven quantity is calculated assuming that quantities combusted andcombustion prices do not change as a result of the rule. While this is not a realistic assumption, it providesa useful way to illustrate the BEQ concept.

Calculating the BEQ is based on a single core economic formula:

Profit = Total Revenues - Total Costs

At breakeven, profit equals zero (i.e., revenues are just large enough to cover all costs):

0 = Total Revenues - Total Costs

Revenues and costs may then be broken into their components:

Total Revenues = [(Price/ton + energy savings/ton) * Tons burned]

Total Costs = [Total fixed costs + (Variable cost/ton * Tons burned)]


We can first define the following variables for the equation:

-- Let $P/ton = price/ton + energy savings/ton-- Let Q = Quantity of waste burned in tons-- Let $FC = total fixed costs-- Let $VC/ton = total variable costs/ton of waste burned

We can then solve the following equation to find the zero profit or breakeven quantity point:

0 = [$P*Q] - [$FC + $VC*Q]

$FC = Q tons ($P-$VC)

Q tons = $FC/($P-$VC) = Breakeven Quantity

Note that in the short-term, $FC includes only the new fixed costs of the rule, such as new pollutioncontrol devices plus baseline fixed O&M costs. All of these fixed costs would be avoided if the facilitychose not to continue burning hazardous waste prior to investing in compliance. In the long-term, thecompany's old equipment will wear out and require replacement. Therefore, $FC for the long-term BEQincludes both the new fixed costs of the rule and the baseline fixed O&M and capital costs.

The data used in the BEQ calculation come from a number of sources. The fixed and variable costsassociated with compliance come from the model plants analysis. Fixed and variable costs in the baselinecome from the baseline cost analysis. As mentioned above, data on prices received by kilns andincinerators are taken from EI Digest. We estimate energy savings based on data in EPA's Cement KilnReport to Congress as well as data gathered under EPA's Combustion Emissions Technical ResourceDocument (CETRED) effort. Specific citations can be found in Section 4.1.1.1, as well as in AppendicesA and E.

Relative to the other basic screens, the BEQ analysis provides a more precise indication of whethera combustion unit is likely to continue burning waste. To assess whether a combustion unit is likely to beable to meet its breakeven quantity, we can compare the BEQ to the quantity of waste currently burned atthe unit. If the BEQ significantly exceeds current tons burned, the unit is likely to cease waste burning.

The BEQ analysis is affected by many of the same uncertainties discussed earlier. Mostsignificantly, inaccuracies in the quantity burned at each unit affect our evaluation of each unit's ability tomeet the BEQ. The lower quantities burned at on-site incinerators will result in more facilities beingunable to meet BEQ (see results discussion below). This result may not be valid if available dataunderstate waste burned quantities. EPA requests additional data from industry on the quantities ofhazardous and non-hazardous waste burned in on-site incinerators.

Greg Kryder and Bruce Springsteen, Energy and Environmental Resources, memorandum to Doug22

Koplow, Industrial Economics, June 27, 1995.

In theory, cement kilns could cover compliance costs associated with hazardous waste burning by23

increasing the prices they charge for cement. Given that cement is a commodity and that many cementkilns do not burn hazardous wastes (and would therefore not incur the compliance costs), kilns are unlikelyto be able to finance compliance by raising cement prices.


In addition, isolation of variable compliance costs is not possible given the structure of the existingcost models. EPA has made the simplifying assumption that all new compliance costs are fixed. This isessentially true for many technologies. For example, the cost associated with improved electrostaticprecipitator (ESP) operation generally will not increase with each ton of waste burned (aside fromincrements to dust treatment and disposal, a minor component of O&M costs). Similarly, the cost ofcarbon injection is more a function of flue gas flow than tons burned. Spray towers, packed towers, andionizing wet scrubbers (IWS) are the only technologies for which the assumption of zero variable costs maybe inaccurate; even with these technologies, the percentage of O&M cost that is variable is estimated tobe below 30 percent for cement kilns and incinerators. Assuming that all new compliance costs are fixed22

leads to some inaccuracies in the estimates of BEQ, although the direction of error is difficult to predict.

4.1.5.2 Refined Breakeven Quantity Analysis

To more accurately reflect decisions that might be faced by combustion facilities, the Agencyrefined the BEQ analysis to incorporate two new contingencies:

! Combustion facilities may be able to cover compliance costs by passing throughportions of the cost increase to generators. This would effectively lower thebreakeven quantity that the unit must meet.23

! Combustion facilities may have the option of consolidating waste burning amongmultiple combustion units at the facility, allowing them to burn waste at someunits while discontinuing hazardous waste combustion at others.

This section discusses our method for modeling these interactions and arriving at more refined estimatesof facility closure, price changes, and waste flow changes. Exhibit 4-4 demonstrates, in flow chart form,the process followed in the refined BEQ analysis.

4.1.5.2.1 Pricing Increases

All combustion facilities that remain in operation will experience increased costs under theproposed regulation. To protect their profits, each will have an incentive to pass these increased costs onto their customers in the form of higher combustion prices. Although generators will not want to pay thehigher prices, they will have to unless they have less expensive waste management alternatives (see Section4.2.2).

In fact, other factors such as transportation costs will affect which facilities are the least expensive to24

particular generators. In addition, the price of combustion will vary by the method of delivery (e.g., bulkversus drum), the form of the waste (e.g., liquid versus solid), and the contamination level (e.g., metals orchlorine content). This makes it more difficult to compare various waste management options.


Exhibit 4-5 (below) illustrates how price pass-through would work in theory. This exhibitillustrates a number of important principles about hazardous waste combustion markets.

! Waste will be sent to the least expensive alternatives first, all else being equal.24

! Both baseline costs of hazardous waste combustion and new compliance costs varysignificantly across combustion units, even within the same sector. Thus,regulatory changes can affect different units in very different ways.

! Prices will rise to the point at which all demand for waste management is met. InExhibit 4-5, the last tons are managed in the non-combustion/waste minimizationalternative at a cost of $230 per ton. This would become the market price.Combustion units A, B, and C would each set their prices at about $230 per ton inorder to maximize their profits. The least efficient management option would earnjust enough to stay in business, but would not recover capital costs. CombustionUnit D would exit the market.

Exhibit 4-5

SIMPLIFIED EXAMPLE OF DETERMINATION OFNEW MARKET PRICE FOR COMBUSTION

Assume 100 Tons n Unit n Unit n Unit Management/ UnitRequire Management A B C Waste Min D

Combustio Combustio Combustio Alternative Combustion

Cost/ton of Waste $145 $175 $220 $230 $240

Tons of capacity 35 25 35 100 300

Remaining tonsrequiring treatment 100-35 = 65 65-25 = 40 40-35 = 5 5-5 = 0 0

The real hazardous waste combustion marketplace is much more complex than the five optionsshown above. Estimating the cost of combustion at which the last ton of waste would be combusted isdifficult due to pricing variations by region, waste stream, and generator. Rather, we have adopted somesimplifying assumptions that should provide a reasonable approximation of these markets:


! We compare the average total cost of combustion by industry sector. Thisincludes both the baseline combustion costs and the new compliance costs. Theindustry sector with the lowest costs is the most-efficient, and will have thegreatest power to pass through compliance costs in the form of higher prices.

! We then calculate the median compliance cost/ton for combustion units in thelowest cost sector after compliance and assume that this amount could be passedto customers in the form of higher prices.

! Using the median compliance costs for the lowest cost sector as the basis for aprice increase is a conservative assumption. As Exhibit 4-5 illustrated, the newmarket price will be set by the most expensive facility that is still needed tomanage the existing supply of waste. The capacity that exists at facilities in theleast cost segment with compliance costs at or below the median for that segmentwill not be sufficient to manage all the wastes requiring combustion. We believethat the actual price increase will most likely be higher than the mediancompliance costs for the lowest cost sector.

! We further assume that the dollar value of this increase would be matched by othercombustion sectors, even if it exceeded their new compliance costs, in an effortto maximize profits. We assume that because the price differential (in dollars)between sectors would be the same as before incurring the new compliance costs,that the market share of each combustion sector would not change.

! Price increases will be capped by the availability of substitutes for combustion(i.e., waste minimization and non-combustion treatment alternatives). Thesealternatives will constrain both price increases by the lowest cost sector, and theability of higher cost sectors to match these increases. Given an absence of gooddata on the price at which these alternatives are viable, we evaluate the impact ofthe proposed rule under both a low and a high price pass-through scenario. Thelow scenario evaluates market impacts if alternatives are available at close to thecurrent market price of combustion.

4.1.5.2.2 Unit Consolidation

In a further attempt to more accurately model industry behavior, we allow for consolidation ofwaste burning units at a facility (illustrated in Exhibit 4-6). The logic behind this is that many hazardouswaste combustion facilities have more than one permitted combustion unit at the same site. Each unit mayburn too little waste to meet the BEQ. However, the facility may be able to move waste between unitsto minimize regulatory impacts. Consolidation offers two benefits to the combustion facility. First, it

A number of facilities report tons burned at the facility, rather than the unit, level. This method of25

reporting hides possible variances in tons burned across the units that would illustrate that some units areabove BEQ and some below. The consolidation routine helps overcome this gap by evaluating how manyunits would need to close in order to bring the remaining units above BEQ.

If large, these waste diversions could help remaining combustion units to meet their BEQ.26

Over the long-term, kilns that remain in business will need both efficient cement production and efficient27

hazardous waste combustion services. Otherwise, plants with both will be able to underprice them.


reduces compliance expenditures because not all units are brought into compliance for hazardous wasteburning. Second, it increases the throughput at the units that do remain in operation. Together, the unitsremaining in operation are more likely to meet their BEQ.25

The consolidation routine closes one unit at multi-unit facilities and distributes the waste from theclosed unit equally among the remaining open units. If the open units still do not meet BEQ, the processis repeated until either the units remaining open meet BEQ or there is only a single unit left. If the singleopen unit does not meet BEQ, we assume that the entire facility will cease burning hazardous wastes, andthe waste will be diverted to other combustion facilities.26

Data limitations regarding unit-level practical capacities create some uncertainty in theconsolidation routine. Practical capacity is the best measure of how much hazardous waste a unit can burnwithout plant modifications. Where particular units have a practical capacity that is less than their BEQ,we may project that they continue to combust hazardous wastes by consolidating wastes from multiple unitswhen this would be impossible in reality due to capacity constraints. Data limitations of this nature aremost prevalent in the cement kiln sector. Examining the units for which we do have unit-level data onpractical capacity suggests that the BEQ after a price increase is generally well below the unit's practicalcapacity for hazardous waste combustion. As a result, we do not expect there to be substantial errors inour market exit estimates from this source of uncertainty.

4.1.5.2.3 Limits to the BEQ Approach

The refined BEQ identifies whether a combustion unit is making money on hazardous wastecombustion or not, and facilitates comparison across units and sectors. However, even units meeting BEQmay experience reduced profitability, if compliance costs cannot be shifted to generators through priceincreases. Thus, units meeting BEQ may decide to stop burning hazardous waste if profits fall too low.

In addition, some marginal cement kilns may subsidize low cement profits with hazardous wastecombustion profits. While such behavior will not be possible over the long term, it may occur in the shortterm. Reduced hazardous waste subsidies to cement production could theoretically make some entire27

plants (joint cement production/hazardous waste combustion) non-viable. Such effects of the rule dependheavily on the operating parameters of specific plants. EPA invites such plants to provide the detailed datanecessary for the Agency to assess the likelihood of closure.

This method assumes that on-site incinerators implicitly earn revenue on waste burning, i.e., burning 2528

percent less waste would require shipping the waste to a commercial combustion facility. Generatorsoperating on-site incinerators also may be able to reduce waste burning by pursuing waste minimization.


4.1.5.2.4 Operating Profits

As noted above, reduced profits from the rule are likely even if units continue to meet BEQ, andtherefore continue to burn hazardous wastes. To evaluate changes in profitability, the Agency hasestimated operating profits by sector in the baseline and under the MACT options. Operating profits arecalculated as follows:

Hazardous Waste Combustion Revenues + Avoided Energy Purchases +(Price Pass-Through if applicable x tons burned) - Baseline Cost ofHazardous Waste Combustion - New Compliance costs, if applicable.

Operating profits are before taxes and overhead charges and provide a good basis of comparisonacross sectors. However, after-tax profits will be lower than operating profits. In the baseline, both pricepass-through and new compliance costs equal zero. Note also that if many units are treating waste burningcapital as sunk it is possible for operating profits to be negative.

4.1.6 Method for Analyzing the Impact of Feed Reductions

As described above, some facilities may choose to comply with the proposed MACT standards byreducing the quantity of pollutants fed to the combustion unit. Therefore, the model plants analysisconsiders the pollution control measures that would be required if all facilities reduced by 25 percent thequantity of metals (including mercury) and chlorine fed to the combustion unit. When these reductions areintroduced, some facilities will require less costly pollution control measures because stack gasconcentrations of certain hazardous air pollutants (HAPs) are reduced.

To gauge the economic impact of these reductions, EPA performed two analyses. First, for eachregulatory option we examined the number of combustion units that shift to a new model plant when a 25percent feed reduction is introduced. These are the only facilities that could potentially have an incentiveto pursue reduced feed since such a change would mean reduced expenditures on pollution control. Thepurpose was to determine whether the feed reduction has a substantial impact on the required pollutioncontrol measures.

Second, we evaluated whether combustion facilities would have an economic incentive to pursuefeed reduction. While feed reduction may limit necessary expenditures on pollution control, it will alsoaffect waste burning revenues. Revenues would be reduced in one of two ways:

! Combustors may choose to simply burn 25 percent less waste, thereby reducingrevenues (or avoided costs in the case of on-site incinerators) proportionately.28

! Combustors may request that blenders supply them with waste that contains 25percent less metals and chlorine. While definitive data do not exist, fuel blendersindicate that there would be a cost associated with supplying this lower-concentration waste (see Section 2.2 for additional information on fuel blenders).Blenders estimated a 50 percent increase in the cost of waste handling, assuming

Personal communication with fuel blenders, including Brian Dawson (Brian Dawson and Associates);29

Bob Campbell (Pollution Control Industries (PCI)); Brad Lamont (Romic Environmental Technologies);Scott Ellis (Cadence); George Anderson (Waste Research); and Chris Goebel (National Association ofChemical Recyclers (NACR)); 22 February, 1995.

40 CFR § 266, Subpart H30

§ 266.108 Small Quantity On-Site Burner Exemption31

EPA investigated alternative approaches for establishing new small quantity burner cutoffs under this32

rule. Efforts were impeded by a lack of data. A cutoff based on the distribution of burn quantities at allwaste burning facilities could not be used because available data from BRS were insufficient for estimatingper-unit (as opposed to per-facility) burn quantities. A risk-based cutoff for waste burning similarly couldnot be used due to the absence of emissions profiles for the small units.


that they can meet demand by segregation and dilution. Blenders will likely pass29

along at least a portion of this cost increase to combustion facilities.

To characterize the incentive that combustion facilities will have to pursue feed reduction, we discussseveral examples demonstrating that, in some cases, the revenue impact of feed reduction outweighs thepollution control equipment savings.

4.1.7 Method for Analyzing Removal of Small Quantity Burner (SQB) Exemption

The Boiler and Industrial Furnace (BIF) Rule, promulgated in 1991, introduced emissions and otheroperating standards designed to control the emissions from, and the waste handling procedures at, waste-burning boilers and industrial furnaces. The rule included an exemption for facilities burning only small30

quantities of hazardous waste. Under the proposed MACT, this exemption would be temporarily removed31

until improved emissions data can be gathered from industrial boilers and other facilities that are likely toclaim the exemption.32

EPA has developed an analysis to characterize the worst-case impact of the removal of the smallquantity burner exemption. The Permits and State Programs Division has compiled a list of the 82 facilitiesthat have filed for the BIF rule exemption. Using Dun & Bradstreet data, we identified the SIC and size(as measured by sales and number of employees) of each facility. To evaluate the economic impact ofremoving the exemption, we first assume that these facilities would pursue off-site management of thewaste. Little is known about the characteristics or specific quantities of waste for each facility. Therefore,we make the following additional assumptions:

Under the exemption, greater quantities are allowed at greater stack heights because of the improved33

dispersion of pollutants experienced with higher stacks. See 40 CFR § 266.108(a)(1).

EI Digest, September, 1994.34

Data were insufficient to calculate BEQs for small quantity burners.35


! The quantity of waste assumed for each facility is the maximum quantity allowedunder the exemption -- 91.2 tons per year. This quantity corresponds to the annualburn quantity for a facility with a stack height of 115 meters or greater.33

! The assumed price of off-site disposal of wastes currently burned by SQBs isbased on reported prices charged by fuel blenders. The inherent assumption is thatSQBs will likely send waste to blenders, not directly to combustion facilities. Forthe lower bound, we assume a per-ton cost equivalent to the price charged byblenders for liquid wastes ($260) while in the upper bound we assume the pricecharged for solids ($1,280).34

To place the estimated off-site waste management costs in context, we compare these costs tofacility revenue where revenue data are available. If costs represent a substantial portion of sales,35

removal of the exemption is more likely to have an effect on the economic viability of the facility. EPArequests additional data from industry to help refine this portion of the analysis prior to final rulemaking.

4.2 RESULTS OF ECONOMIC IMPACT ANALYSIS

4.2.1 Results of Basic Economic Screens

Our basic economic screens are designed to characterize how the proposed MACT standards wouldchange the cost structure of combustion facilities. While this change in cost structure is useful for assessingthe immediate impacts of the rule, it is not sufficient for predicting changes in industry behavior. This isthe objective of the breakeven quantity analysis discussed later in this chapter.

Below, we review the results of the three basic economic impact screens: cost per ton of wasteburned, costs as a percent of revenues, and costs as a percent of baseline burning costs.

4.2.1.1 Cost Per Ton of Waste Burned

Compliance costs per ton of waste burned illustrate the impact of the proposed MACT standardson a standard basis across different plant sizes. Exhibit 4-7 presents compliance costs per ton for the sevenregulatory options. In the commercial sectors, lightweight aggregate kilns bear the greatest cost per tonunder all options. This is a result of both the larger per-unit cost incurred as well as the relatively limitedtonnage burned per unit. The regulatory burden per ton of waste burned is similar for cement kilns and

As we will discuss later, on-site incinerators may be burning both hazardous and non-hazardous waste.36

Even if this is the case, however, the costs of this rule are attributable to the cost of hazardous wasteburning; hence, it is accurate to say that costs per ton of hazardous waste are high.


commercial incinerators (about $100 per ton), except where kilns must install PIC controls (Options 2a and4). Costs per ton in the on-site incinerator sector are extremely high (thousands of dollars per ton), drivenby very low quantities burned at many facilities, which skew the average cost per ton upward.36

Costs vary little across regulatory options. For example, commercial incinerator costs per tonincrease from $77 to $103 per ton moving from the floor to Option 1b, but vary little for other options. Forkilns, costs increase greatly for PIC control options (2 and 4), but otherwise vary minimally. On-siteincinerator costs show the greatest increase between Option 1b to 1c where more stringent mercury controlsare introduced.

4.2.1.2 Costs as a Percentage of Waste Burning Revenues

Compliance costs as a percentage of revenue illustrate how much of current waste burning revenuesmight be absorbed by new compliance costs. Exhibit 4-8 summarizes the comparison of costs and revenuesper combustion unit. The figures represent the percentage of combustion units that fall into each percentiletier. For example, in the floor, 7 percent of all cement kilns have compliance costs that are less than 10percent of the revenues (and energy savings) earned on waste burning.

Looking across sectors, we see that commercial incinerators incur costs that represent a relativelysmall percentage of revenue. Nearly three-quarters of the units incur costs less than ten percent of revenuesfor the floor and Option 1b. Only 18 percent of the units incur costs exceeding 20 percent of waste burningrevenues. This is because, on average, commercial incinerators burn more higher-revenue solids andsludges than do BIFs. In contrast, cement kiln and LWAK impacts are higher because: (1) a number ofkilns burn very little waste, and (2) kilns burn more liquids that earn lower revenues on a per ton basis.More than 60 percent of the LWAKs incur costs exceeding 75 percent of waste burning revenues under allMACT scenarios. On-site incinerator impacts are relatively limited for one-third of the units, despite theextremely high average compliance costs per ton of waste burned. This result is due to a skeweddistribution of waste burning across units, with many units burning very little waste; these units have highcosts relative to revenue, while units with sufficient waste are less affected.

Based on this measure of impacts, increased stringency across regulatory options has little effect.Moving from the floor to Option 1b entails only a small redistribution of cement kilns, on-site incineratorsand commercial incinerators; LWAKs do not change at all.

EXHIBIT 4-7

AVERAGE TOTAL ANNUAL BASELINE COST OF BURNING WASTEAND COMPLIANCE COSTS PER TON OF HAZARDOUS WASTE BURNED

(Before Consolidation)


$28,460$806$194$104Baseline

Compliance Costs

$4,806$77$139$84Original Floor

$4,970$103$139$90Option 1b

$7,545$107$193$112Option 1c

$7,629$113$470$434Option 2a

$7,629$113$470$112Option 2b

$7,545$107$216$112Option 3

$7,626$113$494$434Option 4

Notes:Compliance costs include CEM costs.1.Average compliance costs per ton exclude units currently2.not burning hazardous waste.On-site incinerator baseline and compliance costs per ton3.are high due to the large number of on-site incinerators that reported low tons burned data to BRS in 1991. If facilities are burning larger quantities of hazardous wastecompliance costs per ton would actually be lower. If facilities are burning large volumes of non hazardous waste in addition to the hazardous waste, baseline costs per ton would be lower.

EXHIBIT 4-8

NEW COMPLIANCE COSTS AS A PERCENTAGE OF HAZARDOUS WASTE BURNING REVENUES(percentage of permitted combustion units; see Note 4)

On-site IncineratorsCommercial IncineratorsLWAKsCement Kilns

>75%51-75%21-50%10-20%<10%>75%51-75%21-50%10-20%<10%>75%51-75%21-50%10-20%<10%>75%51-75%21-50%10-20%<10%

28%7%24%7%33%12%6%0%12%71%62%8%23%0%8%20%13%44%16%7%Original Floor

30%9%20%7%33%18%0%0%12%71%62%8%23%0%8%22%13%49%11%4%Option 1b

31%11%19%7%31%18%0%0%18%65%69%8%15%0%8%31%7%49%13%0%Option 1c

31%11%19%15%24%18%0%0%24%59%77%15%8%0%0%91%4%4%0%0%Option 2a

31%11%19%15%24%18%0%0%24%59%77%15%8%0%0%31%7%49%13%0%Option 2b

31%11%19%7%31%18%0%0%18%65%77%0%15%8%0%31%7%49%13%0%Option 3

31%11%19%11%28%18%0%0%24%59%77%15%8%0%0%91%4%4%0%0%Option 4

Notes:Compliance costs include CEM costs.1.Compliance costs as a percent of revenues = [Total compliance costs per ton]/[Waste burning revenues per ton + Energy savings per ton]2.On-site incinerator revenues are equal to the costs generators avoid by not shipping the waste to a commercial incinerator (waste fees charged + 3.transportation costs).High-end of range (>75 percent) includes units not currently burning hazardous waste.4.

Note that this cost increase should not be interpreted as an increase in the cost of cement production.37

Unlike other regulatory impact analyses that treat compliance costs as increases in the cost of producingthe primary product (e.g., plastic, wood pulp), our analysis focuses solely on the cost increase in wasteburning services.


4.2.1.3 Compliance Costs as a Percentage of Baseline Costs

To evaluate the impact of the proposed MACT standards on the cost structure of combustionfacilities, compliance costs as a percentage of baseline burning costs is an illustrative measure. Exhibit 4-9presents a percentile breakdown of this measure. Cost increases for BIFs are substantial. For example,three quarters of all cement kilns have over a 50 percent increase in the cost of burning under Option 1b.This occurs because in the baseline, waste burning at BIFs requires only limited investments in storage andwaste handling equipment, as well as permitting. Baseline costs for BIFs do not include the purchase andoperation of the actual kiln. Therefore, addition of the pollution control equipment needed to comply withthe proposed MACT standards represents a significant cost increase. Baseline costs include existing37

investments in pollution controls driven by existing regulations, but do not include the costs of controls thatmay be required under pending regulations.

Most commercial incinerators will have smaller changes in their basic cost structure (50 to 70percent of the units will see baseline costs rise by less than 10 percent). This is due to baseline costs thatare already relatively high (because they include basic burning equipment such as the kiln itself). Changesin the cost structure for on-site incinerators are also relatively large, although not as large as experiencedby BIFs. Twenty-five to 30 percent of the units will face an increase in costs of more than fifty percent.In addition, one must bear in mind that on-site incinerators may burn non-hazardous wastes in theircombustion units. If this is true, we have overstated the baseline costs of burning, and impacts on the coststructure of on-site incinerators may be greater than reflected here.

As noted in reference to the other measures of economic impact, the variation across regulatoryoptions is limited. Meeting the floor standards entails the greatest addition to baseline costs; increasingstringency for various HAPs has little impact on compliance costs relative to baseline costs. Only whenPIC control standards are introduced (Options 2 and 4) does the distribution of units change drastically.

4.2.2 Alternatives to Combustion

With an increase in costs, combustion facilities will attempt to increase prices charged to generatorsand fuel blenders. Large increases in the price of combustion will induce waste generators to search foralternative waste management methods. These alternatives include waste minimization and a variety ofnon-combustion technologies capable of managing particular waste streams that are now combusted. Ifthese options are available in sufficient capacity, they will increase the elasticity of demand forcombustion. The price at which these alternatives to combustion are available will limit how muchcombustors can increase their prices in the face of new compliance costs.

EXHIBIT 4-9

NEW COMPLIANCE COSTS AS A PERCENTAGE OF BASELINE COSTS OF HAZARDOUS WASTE BURNING(percentage of permitted combustion units; see Note 4)


>75%51-75%21-50%10-20%<10%>75%51-75%21-50%10-20%<10%>75%51-75%21-50%10-20%<10%>75%51-75%21-50%10-20%<10%

4%20%30%24%22%0%0%6%24%71%38%0%38%23%0%33%36%24%2%4%Original Floor

7%20%26%26%20%0%0%12%18%71%38%0%38%23%0%44%31%20%2%2%Option 1b

9%22%26%31%11%0%0%12%24%65%54%23%15%8%0%60%33%7%0%0%Option 1c

13%19%28%31%9%0%0%12%35%53%100%0%0%0%0%100%0%0%0%0%Option 2a

13%19%28%31%9%0%0%12%35%53%100%0%0%0%0%60%33%7%0%0%Option 2b

9%22%26%31%11%0%0%12%24%65%85%0%15%0%0%60%33%7%0%0%Option 3

9%24%28%30%9%0%0%12%35%53%100%0%0%0%0%100%0%0%0%0%Option 4

Notes:Compliance costs include CEM costs.1.Compliance costs as a percent of baseline costs = [Total annual compliance costs/Total annual baseline costs]2.Total annual baseline costs = Annualized fixed capital and fixed operating costs + (Variable operating costs * Hazardous waste burned).3.Percentages include units not currently burning hazardous waste.4.

An increase equal to 25 percent of new compliance costs translates into increases in the average price38

of combustion of between two and 13 percent under all MACT options other than PIC control (see Exhibit4-13).


Available economic data on the cost of these waste minimization and management alternatives andtheir regional distribution are not precise enough for us to pinpoint the maximum price increase thatcombustors could pass through. However, the discussion to follow suggests that alternatives to combustiondo exist for many wastestreams, and may constrain combustion price increases. Therefore, EPA hasconducted a breakeven quantity analysis incorporating three price pass-through scenarios to bound ourevaluation of the impacts of the rule. The low scenario assumes price increases equal to 25 percent of themedian compliance costs in the lowest-cost sector (cement kilns). This scenario would be appropriate ifwaste minimization and alternatives were available at prices only slightly higher than current combustionprices. The high scenario assumes that the entire median compliance cost in the lowest-cost sector (again38

cement kilns) could be passed through in the form of higher prices. The higher scenario is appropriate ifwaste minimization and alternatives are significantly more costly than the current cost of combustion, orif they are not applicable to much of the waste that is now burned. Finally, a zero-percent price pass-through analysis bounds the impacts of the rule by assuming that all new compliance costs must be paidfrom existing profits. Market exit is highest under the zero-percent scenario.

4.2.2.1 Waste Minimization

This analysis of waste minimization potential suggests that generators currently burning wastesmay have a number of options for reducing or eliminating these wastes. The results of the analysis aresummarized in Exhibit 4-10. A more complete discussion of results can be found in Appendix C.

Overall, EPA estimates that up to 633,000 tons of waste -- a significant portion of all combustedwaste -- may be amenable to waste minimization. Three waste generating processes account for most ofthe tonnage reduction. First, a large quantity of combusted waste is generated by solvent and productrecovery/distillation procedures, primarily in the organic chemicals industry. These generators havepracticed recovery/distillation for many years, and therefore may be using outdated equipment that is lessefficient than modern technologies. The most widely applicable improvement that could be introduced isuse of multi-stage stills. This technology is estimated to reduce total generation of waste streams amenableto recovery/distillation by 75 percent and typically entails a relatively modest capital investment of $60,000to $180,000.

EXHIBIT 4-10

SUMMARY OF WASTE MINIMIZATION ANALYSIS

AnnualWasteCombustionReduction

Savings(tons per year)Waste Minimization TechnologyWaste Generating Process

$1,943,0824,620Aqueous detergent systemsCleaning, Rinsing, and Degreasing

Precision cleaning systems$2,208,5195,409(supercritical carbon dioxide)

$106,472178Non-hazardous paint/coatingPainting and Coatings

$79,817133High vol. low pressure/airless painting

$23,247,44822,089Vacuum distillationSolvent and Product Recovery/Distillation

$215,171,340208,235Multi-stage distillation

$111,540,404187,863Recovery systems (engineered for process)Product Processing

$88,734,254122,139Recovery systems (engineered for process)Process Waste Removal and Cleaning

$6,646,5366,928Recovery systems (engineered for process)Waste and Spent Material Removal

$1,974,2382,935Quality control, extended shelf life, reuseDiscarding and Decommissioning

$68,590,71272,651Recovery and reuse of waste productsPollution Control and Wastewater Treatment

$520,242,822633,180Total =

Source: Jerome Strauss and Peter Von Szilassy, Versar Incorporated, "Preliminary Assessment of WasteMinimization Potential for Combusted Wastes," prepared for U.S. EPA, March 30, 1995.

Note that EPA implicitly assumes that the wastestream managed by the recovery system is primarily39

product that can be reused, thereby allowing large (75 percent) reductions in waste quantity. In contrast,if the waste generated in product processing is primarily low concentration aqueous waste, productrecovery may be high, but the tonnage reduction for the overall wastestream will be more limited. Becausethe Agency has not factored in the form of the waste, it is likely that tonnage reductions are somewhatoverstated.


Waste minimization opportunities may also exist for product processing wastes. Productprocessing wastes are generated from product rinsing, filtering, extraction, and forming. They usuallyconsist of virgin product material mixed with a solvent or water as well as some solids. The primary wasteminimization opportunities entail recovering product for reuse as well as reusing the rinsewaters to themaximum extent possible. Typically, a recovery system entailing a combination of gravity settling, variousstages of filtration, and membrane technology will cost from $80,000 to $300,000, depending on capacityand degree of automation. Such a recovery system will achieve waste reductions of 75 percent or higher.39

As shown, over 187,000 tons could potentially be eliminated by this measure. Most of these reductionswould occur in the organic chemicals industry.

A third major category of waste potentially amenable to waste minimization is generated byprocess waste removal and cleaning activities. The wastes generated usually are comprised of product andintermediates mixed with solid contaminants and water or solvent depending on the cleaning methods. Formulti-product process lines, the waste is generally virgin product that must be removed prior to productchange-over. The technologies and associated costs for waste minimization are similar to those of theproduct processing group; however, product recovery is usually much less due to higher concentrations ofcontaminants. As shown, over 122,000 tons of waste reduction could potentially be achieved, with mostof this occurring in the organic chemicals and pesticides industries.

This analysis is subject to several key caveats that influence the waste minimization potentialreflected in the preceding discussion.

! First, as with our investigation of waste management alternatives, a number oftechnologies appear to have costs low enough that one would expect them to beimplemented already. However, EPA has not evaluated the total costs andbenefits associated with the waste minimization measures. For a wasteminimization measure to truly be cost effective immediately, the total per ton costof reducing the waste must be less than the marginal cost per ton of combustion.The payback method of identifying waste minimization options relies onsimplified capital cost information that understates the total costs by notincorporating operating costs associated with the option. Conversely, the absenceof cost savings from reduced raw material, energy, and labor requirements lead tooverly high estimated costs of waste minimization.

! Second, a variety of obstacles could impede the adoption of waste minimization,including lack of information and lack of capital for new investments (for smallergenerators).


! Third, it is important to reiterate that the analysis considers wastes characterizedat a very general level -- BRS source code and industry. At a greater level ofdetail, there may be characteristics of any given wastestream that preclude orenhance use of the waste minimization measure identified in the analysis.

! Finally, EPA is requesting comments on a number of additional provisions thatcould alter the economic viability of waste minimization options presented here.The possible requirement that on-site combustors better characterize their wastestreams and target waste minimization plans based on this characterization couldencourage additional waste minimization and/or pollution prevention. Similarly,in return for specific waste minimization commitments, the Agency may allowcombustors additional time to comply with the new regulations or delay calling-inexisting permits. While such exceptions will be determined on a case-by-casebasis, they could increase the attractiveness of particular waste minimizationoptions.

These caveats suggest that the specific figures for tonnage reduction presented here are fairlyrough. Nonetheless, the analysis suggests that a substantial portion of combusted waste could potentiallybe reduced through the application of waste minimization approaches. In particular, EPA believes thatthese impacts are most likely to occur in the on-site incinerator sector, where much of the waste amenableto waste minimization is generated. On-site facilities are the most likely to select alternatives to installingnew pollution controls at their facilities because of the relatively small tonnage burned at many of them.Some of these facilities, perhaps even many given the results discussed above, may find that the costs ofinstalling pollution prevention technologies are now less expensive than the prices they would have to payfor commercial combustion of their waste. These added incentives for waste minimization are expectedto increase the adoption of new management approaches in this sector of the combustion market. EPA isrequesting additional information from industry on the likelihood that the new MACT standards will in factresult in greater reliance of waste minimization approaches.

4.2.2.2 Waste Management Alternatives

To evaluate waste management alternatives to combustion, EPA identified the major wastecategories in the BRS, determined which might be amenable to alternative treatment, and then quantifiedthe tonnages in these waste categories. The Agency then assembled available cost information for eachof the feasible alternatives.

The analysis of waste management alternatives suggests that, based on the waste descriptions inthe BRS data, waste management alternatives might exist for a significant portion of waste that is currentlycombusted. In terms of quantity, the largest wastestreams that may be manageable by other means are low-concentration aqueous wastes and solvent-based wastes. Although there are many technologies underdevelopment, this report highlights several that are technologically well-established and are effective evenwhen metals are present (though not all of them remove the metals). The five alternative treatmenttechnologies most applicable to waste streams currently combusted are shown below in Exhibit 4-11. Theyinclude:


Exhibit 4-11

APPLICABILITY OF ALTERNATIVE COMBUSTION TECHNOLOGIES

Air Gravity SolventStripping Distillation Separation Ozonation Extraction

General Location of Unit On-site Off-site On-site off-site (soils) On-siteOn-site (liquids);

Materials Treated:Liquids? X (non-aqueous) X X XSolids/Sludges? X XSoils? X X

Specific Wastes 3,4 3,4 3,4 1,2,3,4,5 3,4,5

Function Separation Separation Separation Destruction Separation

Effective If Waste Contains Yes Yes Yes Yes YesMetals?

Treats Volatile Organics? Yes Yes Yes Yes Yes1

Treats Semi-Volatiles? No Yes Yes Yes Yes1

Description of any Any semi- Any metals, (1) Water saturated Residual Metals (unless acidsHazardous Constituents in volatiles or certain with all hazardous hazardous are used as solvent)Residuals (e.g., residual metals hazardous organics organics and allmetals requiring organics (2) Water with metalsstabilization) dissolved metals

Minimum Economic Scale 200 None (can be 200 2000 (liquids) 2000 (liquids)(tons/year) batch operated) unknown 10,000 (solids, soils,

2 2 3

(solids) sludges)

3

4

Other Limitations? Only for Only separates Very Not effective forstreams of immiscible phases questionable many compounds

dilute volatiles, (e.g., oil and water), applicability fore.g., less than not specific soils; not

100 ppm components effective onchlorinatedcompounds

Key to Waste Streams: (1) inorganics without heavy metals; (2) inorganics with heavy metals; (3) organics without heavy metals; (4) organics with heavy metals; and (5) oil/water

Notes: If not miscible with aqueous portion of stream1

0.1 gallon/minute (gpm) 526 x 10 minutes/yr.2 5

Use activated carbon for smaller volumes.3

Less throughput would probably be thermally desorbed or incinerated.4

Source: Jerome Strauss, Versar, Inc., "Evaluation of Alternative Treatment Technologies for Wastes That Are Currently Being Sent toCombustion Facilities (Revised Draft)," memorandum prepared for Lisa Harris, U.S. EPA and Bob Black, Industrial Economics, Inc., March 30, 1995; and subsequent conversations.

1991 Biennial Report, Draft Data Summaries, 9/29/94. 40


! Air Stripping. Uses temperature, pressure, and other parameters to transfervolatile contaminants from water to air. Off-gases must be captured and destroyedto prevent release.

! Distillation. Uses heat to separate volatile organics from mixed contaminants.Common distillation techniques include batch distillation, fractionation, steamstripping, and thin film evaporation. Off-gases must be captured and destroyed toprevent release.

! Gravity Separation. Uses gravity to separate aqueous wastes having differentspecific gravities (e.g., oil in water).

! Ozonation. Uses dissolved ozone and/or hydrogen peroxide to break downorganic compounds in waste.

! Solvent Extraction. Waste mixed with a solvent that dissolves the organicconstituent of concern. The dissolved waste is then removed from the solvent bydistillation.

Additional information on these alternative treatment methods is presented in Exhibit 4-11. Thetechnologies are applicable primarily to organic waste streams, although ozonation can be used on streamscontaining organics as well. Treatment units tend to be built on-site, rather than as independent commercialunits. As a result, to justify such construction, a facility must generate enough waste to adequately utilizethe unit.

While all of the technologies can be used to treat liquid waste streams, applicability to soils andsludges is more limited. Furthermore, only ozonation is a destruction technology; all of the others merelyseparate the waste constituents for further management (that could involve waste combustion). Each alsohas limitations regarding the types of waste streams it can treat and the residuals that are generated. Theselimitations could disqualify their use in specific situations.

4.2.2.3 Impact of Alternative Treatment on Quantities Combusted

EPA's analysis of alternative management suggests that up to 1.5 million tons of wastes currentlycombusted may be amenable to commercially-available alternative management approaches (Exhibit 4-12).This comprises over 47 percent of the 3.0 million tons burned in the combustion facilities covered by theproposed MACT standards. 40

Were this quantity of waste actually to move out of the combustion sector, the implications forcombustion facilities would be extreme. Many combustion units would no longer burn enough wastes toadequately spread their fixed costs, falling below their BEQ. The impacts would be most pronounced foron-site incinerators because 85 percent of the nearly 1.5 million tons that could be diverted are currently

EPA analysis of the 1991 BRS. Note that the source and form code information needed to conduct the41

alternatives analysis is most readily available for wastes burned at on-site incinerators, and is oftenunknown for wastes burned at other facility types. Therefore, although other facility types may burn wastesthat are amenable to alternative management, the lack of form and source code information prevents theAgency from analyzing this potential.


Exhibit 4-12

QUANTITY OF KEY WASTESTREAMS THAT COULD POTENTIALLY BE MANAGEDBY EACH OF THE CURRENTLY VIABLE ALTERNATIVE TECHNOLOGIES

Wastestream (Form and Source) Technologies Managed (Tons)Applicable Alternative Quantity That Could Be

Wastewaters and Aqueous Wastes (PF06) from Process Waste 275,140Removal and Cleaning (S07) 1,2,3,4,5

Wastewaters and Aqueous Wastes (PF06) From Solvent & 247,428Product/ Recovery/ Distillation (S05) 1,2,3,5

Wastewaters and Aqueous Wastes (PF06) From Product 229,036Processing (S06) 1,2,3,5

Nonhalogenated Solvents & Other Organic Liquids (PF02) From 222,329Process Waste Removal & Cleaning (S07) 2,3,4

Nonhalogenated Solvents & Other Organic Liquids (PF02) From 219,892Solvent & Product/Recovery/Distillation (S05) 4

Still Bottoms (PF03) From Pollution Control & Wastewater 180,282Treatment (S13) 2,3,4

Halogenated Organics/ Solvents (PF01) from Pollution Control & 98,939Wastewater Treatment (S13) 1,2,3,4,5

Total Waste that Could Potentially be Diverted 1,473,046

Codes for Applicable Alternative Technologies: (1) Air Stripping; (2) Distillation; (3) Gravity Separation; (4) Solvent Extraction;(5) Ozonation.

Source: Jerome Strauss, Versar, Inc., "Evaluation of Alternative Treatment Technologies for Wastes That Are Currently Being Sent toCombustion Facilities (Revised Draft)," memorandum prepared for Lisa Harris, U.S. EPA and Bob Black, Industrial Economics, Inc., March 30, 1995.

burned at the on-site facilities. Fourteen percent is currently burned in cement kilns, and less than onepercent is currently burned in commercial incinerators.41

Such a large diversion of wastes is highly unlikely for a number of reasons. First, severalalternatives appear quite inexpensive relative to current combustion prices, yet generators have not stoppedburning their wastes. Second, generators used combustion as a management tool even when the cost perton was significantly higher than today. These factors suggest that the costs of the alternative technologiesare higher than estimated here, and/or that their applicability to the waste streams of concern is limited byother factors. The main factors that seem to be limiting the use of alternatives in the current marketplaceare presented below:

! Cost Estimates for Alternatives May be Understated. Data on the cost of thesealternatives are incomplete (see Exhibit 4-13). The main database that tracks thecosts of these alternative technologies (ATTIC) contains voluntarily-reported costdata from vendors. Even where data are available, there is little information onwhat costs are included in the reported numbers. As a result, there is a stronglikelihood that many of the costs reported do not necessarily represent the fullannualized cost per gallon of treatment.


Exhibit 4-13

COST DATA ON MOST-APPLICABLE ALTERNATIVE TECHNOLOGIES

Air Stripping Distillation Gravity Separation Ozonation Solvent Extraction

Initial Capital Cost -- $1,050,000 -- $70,000 - $260,000 $1,030,000[3] [1]

$130,000 - $160,000 $3,300,000[2]

[1]

[3]

Estimated Capital Equipment Life ~ 5 yrs. ~ 10 yrs. ~ 15 yrs. ~ 10 yrs. ~ 10 yrs.(Years)

Estimated Annual Gallons Equipment -- 35,280,000 *** -- 65,700,000 * 190,008,000 **Can Handle

[3] [3] [3]

Permitting Cost -- -- -- --

Fixed O&M Costs:Soils/Sludges ($/ton) $70 - $380 $27 - $207 $30-$175 $120-$450Liquids ($/1000 gal) $0.29 - $0.66 $0.25-$17 $18Liquids (converted to $/ton) $0.07 - $0.16 $0.06 - $4.08 $4.32

[1]

[1] [1] [1]

[1]

[1]

[1]

Variable O&M Costs Liquids ($/1000 gal) -- $0.01 -- -- --Liquids (converted to $/ton)

[3]

Data Source(s) [1] - ATTIC [1] - ATTIC [1] - ATTIC [1] - ATTIC [1] - ATTIC[3] - Standard [2] - Pollution [3] - StandardHandbook of Prevention October 94 Handbook of

Hazardous Waste Hazardous WasteTreatment and Treatment andDisposal, 1995 Disposal, 1995

Notes: * Annual capacity determination based on a remediation operation; assuming 24 hrs/d times 365 d/y times 7,500 gph (125 gpm) flowrate = 65,700,000 g/yr.** Annual capacity determined based on 728,000 gals/day capacity times 261 days/year = 190,000,000 gal/year.*** Annual capacity determined based on 8,400 h/yr operation times 4,200 gph = 35,280,000 g/yr.

Source of Compiled Data: Jerome Strauss, Versar, Inc., "Evaluation of Alternative Treatment Technologies for Wastes That Are Currently Being Sent toCombustion Facilities (Revised Draft)," memorandum prepared for Lisa Harris, U.S. EPA and Bob Black, Industrial Economics, Inc., March 30, 1995; andsubsequent conversations.

We would expect liability concerns to create a combustion premium for generators. That is, a generator42

will not switch from incineration as soon as another technology is less expensive. However, at some pointthe cost differential between combustion and alternatives would grow toolarge, and even liability-conscious generators would begin shifting from combustion.


! Average Combustion Prices Don't Adequately Reflect Certain High-ValueSegments Such as Solvents. While average combustion prices are well above theestimated price of alternative technologies, actual prices for high-Btu solventwastes may be very low. For example, cement kilns and blenders may accept thewaste for free or even pay generators for the waste. Thus, alternatives for high-Btu solvents (generally separation technologies) may not, in reality, be lessexpensive than combustion (that recovers the Btu-value of the waste).

! Alternative Management Options May Not Adequately Treat the SpecificWaste Streams Now Combusted. This analysis considers waste characterizedat a very general level -- BRS source and form code. At a greater level of detail,there may be characteristics of any given waste stream that preclude use of thetechnology identified in the analysis. For example, two of the inexpensivetechnologies are limited for certain waste characteristics. Air stripping is effectivein removing volatiles but not semi-volatiles. Ozonation is not effective withchlorinated compounds. BRS data do not provide information at this level ofdetail.

! Generators Have Insufficient Quantities to Justify Installation of SpecializedRecovery Equipment. While incineration can handle a wide variety ofwastestreams, some of the alternatives may be more narrow in their applicability.Generators may not have sufficient tonnage to meet the minimum economic scalerequirements.

! Many On-site Units May Consider Incinerator Capital "Sunk". Accordingto BRS data, much of the combustion tonnage for which alternative treatmentexists is currently managed in on-site incinerators. Since the units have alreadybeen built, waste management alternatives (fixed plus variable costs) would needto be less expensive than the variable costs at the on-site incinerator. In suchcases, the generator may switch to an alternative once the incinerator needs to bereplaced, but not earlier.

! Liability Concerns Lead Generators to Favor Combustion. Most of theapplicable technologies are separation, not destruction technologies. Thus, whilethe incineration of low-concentration (low-Btu); high water content wastes makeslittle sense, the definitive destruction ability of combustion is attractive for manyliability-conscious generators.42

The BRS data base does include a major category of waste of "unknown" origin. These may also be43

amenable to waste minimization, although without further data on the source this cannot be analyzed.


4.2.2.4 Summary

This waste minimization and alternative treatments analysis suggests that, over time, alternativesto combustion are likely to erode combustion's market share. Available cost data suggest that generatorscould already realize savings by adopting alternatives. Paradoxically, generators continue to rely oncombustion rather than the alternatives. While it appears likely that the factors described above willcontinue to keep many wastes in the combustion sector in the face of moderate price increases, data on thecost and applicability of these alternatives are not precise enough to assume with any certainty thatcombustors could increase prices significantly in order to cover their new compliance costs. For thisreason, the Agency has evaluated the impacts of the rule assuming no compliance costs could be passedthrough, as well as other, more likely, scenarios. Over the longer-term, however, the shift to thesealternatives could be more significant. In the case where generators treat their on-site incinerator capitalas sunk, large required reinvestments in the units could lead plant managers to reevaluate non-combustionalternatives. Given the limited waste quantities from those on-site units that we expect to stop burninghazardous wastes in the face of the new regulation, the short-term effect on the overall market will besmall. Nonetheless, large increases in the cost of combustion overall will lead to increased interest inalternatives, and some shifts are likely.

As noted previously, evaluation of the BRS data suggest that on-site incinerators have the greatestwaste quantities amenable to waste minimization. As the analysis discussed in the next section indicates,43

these on-site facilities are also the most likely to cease burning hazardous waste as a result of the newMACT standards. Facilities that cannot afford to install pollution controls that comply with the newstandards must face the choice of either sending their waste to a commercial combustion facility oradopting some alternative approach such as waste minimization. Because the costs of waste minimizationare often lower than commercial combustion prices for these industries based on the analysis discussedabove, many are likely to find that pollution prevention is their preferred alternative.

4.2.3 Breakeven Quantity Analysis and Changes in Waste Burning Behavior

The breakeven quantity analysis examines regulatory impacts at the unit level. By comparing thecurrent tons burned with the tons a combustion unit needs to burn to cover costs, the BEQ analysisidentifies combustion units that require increases in the prices they receive and/or the quantities they burnto continue operating profitably.

EPA evaluated a variety of BEQ measures, each of which provides answers to a different set ofquestions:

! Price Pass-Through Scenarios. As discussed, available data do not allow EPAto identify precisely the price at which waste minimization and alternativemanagement methods would begin to divert significant waste quantities fromcombustion. Therefore, the Agency calculates whether combustion units meettheir BEQs under both small price increases and large increases, as well asassuming no price increases. Higher prices reduce the BEQ; as a result, the morethat prices are allowed to rise, the fewer units stop burning wastes. Stated


differently, more inelastic demand for combustion allows combustion facilities topass through costs.

! Short-term Versus Long-Term BEQ. For each BEQ measure, the Agencyevaluates both the short-term and the long-term. The long-term measure capturesbehavior over the capital cycle of the combustion units; some may be able tocontinue operating in the near term, but do not earn enough to replace their wasteburning capital when it wears out.

All measures incorporate waste consolidation prior to evaluating BEQ. Thus, a facility with threeunits permitted to burn hazardous wastes can consolidate waste burning into two or one units in order tomeet the BEQ.

4.2.3.1 Low Price Pass-Through Scenario

Under the low price pass-through scenario, prices are assumed to increase in every sector by anamount equal to 25 percent of the median new compliance costs for the lowest-cost sector. Since totalcosts per ton combusted are lowest in the cement kiln sector for all proposed MACT options, the priceincrease equals 25 percent of the median cement kiln compliance costs. Under most scenarios, thistranslates to an increase of less than $25 per ton (see Exhibit 4-14). On a percentage basis, this priceincrease is minor (less than 4 percent) for incinerators where current weighted average prices are about$600 per ton. The increase is more substantial for kilns (about 12 percent), where current prices are about$200 per ton.

4.2.3.1.1 Short-Term BEQ

As shown in Exhibit 4-15, compliance costs from the proposed MACT would prevent a significantfraction of LWAKs and on-site incinerators from meeting BEQ, but would have a lesser effect on cementkilns and commercial incinerators. Between 20 and 25 percent of the cement kilns are below BEQ, andwould be expected to stop burning hazardous wastes. As nine percent of the permitted kilns do notcurrently burn hazardous wastes, approximately 10 to 15 percent of the market exit may be properlyattributed to this rule. More than half of the units would stop burning hazardous wastes if PIC controlswere required. In contrast, over 80 percent of the commercial incineration units would meet BEQ underall options. Of the 18 percent expected to close, one-third do not currently burn any hazardous wastes andwould likely close even in the absence of this rule. Market exit in this sector is unaffected by the proposedMACT option chosen.

In the LWAK sector, nearly two-thirds of the units do not meet their short-term BEQ in the lowprice pass-through scenario. However, 31 percent of the LWAKs are not currently burning any wastes.Therefore, at most only half of the expected market exit can be properly attributed to the proposed rule.

EXHIBIT 4-14

WEIGHTED AVERAGE COMBUSTION PRICE PER TON ANDINCREASE IN PRICES DUE TO ASSUMED PRICE PASS THROUGH

25%Price pass through assumed: (percentage of median compliance costs for the most efficient sector)

CommercialSector with

LowestOn-siteCommercialLWACementTotal cost/tonIncineratorsIncineratorsKilnsKilnsOptions

Current weightedcement kilns$580$646$188$178average price

Increase in price due to compliance costs passed through

cement kilns$17$17$17$17Original Floor

cement kilns$17$17$17$17Option 1b

cement kilns$22$22$22$22Option 1c

cement kilns$95$95$95$95Option 2a


cement kilns$22$22$22$22Option 3


Notes:Compliance costs include CEM costs.1.Median compliance costs per ton exclude units currently2.not burning hazardous waste.The commercial sector with the lowest total cost per ton (baseline 3.+ compliance cost) drives the assumed increase in combustion prices.Prices for on-site incinerators reflect the cost per ton of off-site treatment that4.generators avoid by burning the waste on-site.Weighted average price per ton = (solids percentage of total waste burned in each5.sector x solids price) + (liquids percentage of total waste burned in each sector xliquids price).

EXHIBIT 4-15

PERCENTAGE OF COMBUSTION UNITS MEETING SHORT TERM BEQ AFTER CONSOLIDATION(Percentage of combustion units; includes units not burning waste in the baseline)



>20% below<20% belowAbove>20% below<20% belowAbove>20% below<20% belowAbove>20% below<20% belowAbove

37%4%59%18%0%82%62%0%38%20%0%80%Original Floor

39%6%56%18%0%82%62%0%38%20%0%80%Option 1b

41%6%54%18%0%82%62%0%38%24%0%76%Option 1c

30%7%63%18%0%82%62%0%38%51%0%49%Option 2a

41%6%54%18%0%82%69%8%23%24%0%76%Option 2b

41%6%54%18%0%82%62%0%38%24%0%76%Option 3

30%7%63%18%0%82%62%8%31%51%0%49%Option 4

Notes:Compliance costs include CEM costs.1.Percent of units currently not burning waste:2.

9% Cement Kilns31% LWAKs

6% Commercial Incinerators11% On-site Incinerators (burning less than 50 tons)

Since some units will stop burning hazardous wastes at facilities that continue to burn wastes at other44

units, facility exit will be lower than unit-level exit.


On-site incinerators are roughly split between units meeting short-term BEQ and those that do not. Thisreflects the wide spectrum of waste quantities burned at these units, from a few tons per year up to manythousands. More than 10 percent of the on-sites currently burn less than 50 tons of waste per yearaccording to BRS data, comprising roughly 20 percent of the on-site units that do not meet their BEQ. Theon-site incinerator results show that more units meet their BEQ in the more expensive options for cementkilns and LWAKs than at the floor. This outcome reflects the fact that as commercial incineration pricesrise, the ability to avoid having to ship wastes off-site becomes more valuable. Thus, more on-site unitsare able to meet their BEQ.

4.2.3.1.2 Long-Term BEQ

The long-term BEQ measure is most interesting when compared to the short-term BEQ resultsshown above. Exhibit 4-16 presents the percentage of units in each sector that do not meet their long-termBEQ under the low price pass-through scenario. In most sectors, fewer combustion units meet their long-term BEQ than their short-term BEQ. This reflects the fact that long-term viability requires recovery ofall costs of combustion, including baseline waste burning capital.

The difference between short-term BEQ and long-term BEQ is larger in the incinerator sectorsbecause baseline capital costs are higher in these sectors than for BIFs. Commercial incinerator exitincreases from 18 to 29 percent, while on-site incinerator exit rises by about 15 percentage points undermost proposed MACT options. Both LWAKs and cement kilns show only minor differences between theshort-term and long-term BEQs.

As with the short-term BEQ, the choice of the proposed MACT option has little impact on the rateof exit from the hazardous waste combustion market. The exceptions are the options requiring PIC controlsat the kilns, which demonstrate markedly higher rates of market exit in both the short- and the long-term.

4.2.3.1.3 Expected Cessation of Hazardous Waste Combustion at the Facility Level

The breakeven quantity analysis provides the information needed to assess whether a particularcombustion unit is viable. However, evaluating market dislocations must also incorporate facility-levelimpacts. The BEQ analysis feeds directly into this evaluation; where no unit at a facility can meet BEQ(even after wastes are consolidated), we assume the facility will cease burning hazardous wastecompletely.44

Facility-level impacts provide the best measure of regional economic dislocations. A cement plantthat consolidates hazardous waste burning in two units on site rather than the previous three will generatesmaller economic impacts than if the plant stops burning wastes altogether. One important caveat is that,for most sectors, exiting the hazardous waste combustion market is fundamentally different from closinga plant. Cement kilns or LWAKs that stop burning hazardous fuels do not stop making cement andaggregate. On-site incinerators are generally located at large industrial facilities such as chemical plantsor refineries. Production is likely to continue even if the wastes are sent off-site for management. Onlyin the case of a commercial incinerator would exit from hazardous waste combustion markets signal theactual closure of the plant.

EXHIBIT 4-16

PERCENTAGE OF COMBUSTION UNITS MEETING LONG TERM BEQ AFTER CONSOLIDATION(Percentage of combustion units; includes units not burning waste in the baseline)




54%6%41%29%0%71%62%0%38%24%2%73%Original Floor

54%6%41%29%0%71%62%0%38%24%2%73%Option 1b

54%7%39%29%0%71%62%0%38%29%2%69%Option 1c

52%2%46%29%0%71%62%8%31%51%2%47%Option 2a

56%6%39%29%0%71%85%0%15%29%2%69%Option 2b

54%7%39%29%0%71%62%8%31%29%2%69%Option 3

52%2%46%29%0%71%62%15%23%51%2%47%Option 4





Exhibit 4-17 summarizes market exit by sector. We estimate that two cement-making facilitiescurrently burning hazardous wastes will stop doing so as a result of the proposed rule (though this jumpsto 9 facilities if PIC controls are required). To avoid attributing market exit of facilities not viable evenin the baseline, Exhibit 4-16 excludes combustors that currently burn no (or less than 50 tons per year inthe case of on-site incinerators) hazardous waste even though permitted to do so. Under all options, weestimate four commercial incinerators will exit, versus one LWAK under most options. As with cementkilns, market exit rises to as high as four facilities if PIC controls are required. On-site incinerator exit isbetween 44 and 61 facilities.

Again, comparing short-term and long-term exit at the facility level provides some interestinginsights (see Exhibit 4-18). Exit increases minimally at BIFs because their relatively low baseline capitalcosts mean that there is little difference between the short run and long run costs that must be recovered.Likewise, the difference between the long- and short-run is moderate for commercial incinerators becausethey generally have a substantial quantity of waste over which to spread capital costs. In contrast, closuresfor on-site incinerators nearly double under many MACT options because of substantial baseline capitalcosts and sometimes small waste quantities over which to spread the costs.


Compliance costs reduce operating profits in two ways. First, some firms stop burning, therebyeliminating any waste burning profits they might have had. Second, the increased costs may not becompletely recovered through price increases, depressing the profitability of hazardous waste combustionactivities at facilities that continue to operate. Under the low price pass-through scenario (see Exhibit 4-19), average operating profits decline in all sectors under most options as new compliance costs must beabsorbed by profits. Under all MACT options, operating profits per ton fall substantially more in thecement kiln, LWAK, and on-site incinerator sectors than in the commercial incinerator sector. Declinesin the operating profits of the BIFs are driven by relatively high compliance costs. Although compliancecosts are somewhat lower for the on-site incinerator sector, the tonnage over which to spread these costsis generally lower, driving up the estimates of compliance cost per ton. The commercial incinerators haveboth relatively low compliance costs and high tonnage, and therefore show small declines in profitabilityeven in the low price pass-through scenario. Since declines are smaller than for cement kilns, the rulewould be expected to improve the competitive position of incinerators somewhat.

Although many sectors experience a significant percentage decline in their operating profitmargins, profit margins even after the rule for both cement kilns and BIFs remain quite strong -- betweentwo and three times those at commercial incinerators. Thus, even with recovery of only 25 percent of thecompliance costs incurred, hazardous waste combustion remains profitable on an operating basis for thesesectors. Under the zero-percent price pass-through bounding scenario (see Appendix G, Exhibit G-5),operating profit margins after the rule for all options not requiring PIC controls are very similar to thoseunder the 25 percent pass-through. This again highlights the greater effect that capacity utilization has onperformance relative to price pass-through.

EXHIBIT 4-17

NUMBER OF COMBUSTION FACILITIES LIKELY TO STOP BURNING HAZARDOUS WASTE IN THE SHORT TERM(net of facilities not currently burning waste)


On-siteCommercialCementIncineratorsIncineratorsLWAKsKilns

Facilities currently not14212burning waste (Note 2)

Incremental Facilities Likely to Stop Burning Waste

51412Original Floor

58412Option 1b

61412Option 1c

44419Option 2a

61442Option 2b

61412Option 3

44429Option 4

Notes:Compliance costs include CEM costs.1.Percentage of facilities currently not burning waste in the cement kiln, LWAK, and2.commercial incinerator sector; or burning less than 50 tons per year in the on-siteincinerator sectors. Some additional units may be nonviable in the baseline, leadingus to overestimate closures due to the proposed MACT.

EXHIBIT 4-18

NUMBER OF COMBUSTION FACILITIES LIKELY TO STOP BURNING HAZARDOUS WASTE IN THE LONG TERM

(net of facilities not currently burning waste)





82613Original Floor

82613Option 1b

85613Option 1c

726210Option 2a

85643Option 2b

85623Option 3

726410Option 4

Notes:Compliance costs include CEM costs.1.Number of facilities currently not burning waste in the cement kiln,2.LWAK, and commercial incinerator sectors; or burning less than 50tons per year in the on-site incinerator sector. Some additional unitsmay be nonviable in the baseline, leading us to overestimateclosures due to the proposed MACT.

EXHIBIT 4-19

CHANGE IN AVERAGE OPERATING PROFITS PER TON OF HAZARDOUS WASTE BURNED FROM THE PROPOSED MACT


On-site IncineratorsCommercial IncineratorsLWA KilnsCement Kilns% Margin afterOperating Profit Margin% Margin afterOperating Profit Margin% Margin afterOperating Profit Margin% Margin afterOperating Profit Margin

the Rule% Change$ Changethe Rule% Change$ Changethe Rule% Change$ Changethe Rule% Change$ ChangeOptions

29%-30%($56)19%-14%($17)60%-21%($32)49%-31%($35)Original Floor

33%-25%($48)19%-12%($15)61%-20%($29)44%-39%($45)Option 1b

33%-26%($51)19%-16%($19)62%-22%($35)44%-41%($50)Option 1c

16%-48%($68)21%-6%$227%-66%($122)19%-78%($127)Option 2a

32%-29%($58)18%-19%($24)38%-54%($158)44%-41%($50)Option 2b

33%-26%($51)19%-16%($19)54%-32%($56)44%-41%($50)Option 3

17%-46%($64)21%-6%$229%-64%($141)19%-78%($127)Option 4

Notes:Compliance costs include CEM costs.1.Operating Profits = (weighted average price per ton + weighted average energy savings per ton + assumed price increase due to compliance costs passed through) - (average baseline costs2.per ton + average total annual compliance cost per ton). Assumed price pass-through is a set percentage (shown at the top of this exhibit) of the median compliance cost for the mostefficient combustion sector. As this is a static model, we have capped the price pass-through using the combustion units expected to remain burning hazardous waste even though theoriginal pass-through value included some units expected to stop burning. This is a better approximation of the impetus combustors have to raise prices, though it is not a precise predictor. To address uncertainty regarding the amount prices will rise, a variety of price increase scenarios were used. All other averages were calculated after consolidation, and include only thoseunits that continue to burn hazardous waste.Operating profits exclude overhead, other administrative costs, and taxes. Actual after-tax profits will be lower.3.Percentage Operating Profit Margin = average operating profits per ton / (weighted average price per ton + assumed price increase due to compliance costs passed through). Percentage profit margin after the rule is4.calculated using the same formula with post-rule operating profits and prices.Change in operating profits per ton = Post-rule operating profits per ton - baseline operating profits per ton. Percentage change in operating profits margin = (post-rule operating profits margin - baseline5.operating profits margin) / baseline operating profits margin. Baseline operating profit margins for units remaining open after consolidation can be calculated by dividing the percentage profit margin afterthe rule by one plus the percentage change in the operating profit margin. For consistency, baseline values have been calculated using the median compliance cost per ton for facilities that remain inoperation after the rule for each MACT option.

A direct comparison ignores the fact that commercial incinerators can burn wastes cement kilns cannot,45

and that incinerators handle more higher-value solids and sludges than do kilns.


4.2.3.1.5 Waste Diversions

As units close, wastes are consolidated at the same facility in the remaining units. However, if afacility closes, the wastes must be managed elsewhere. Waste diversion levels are important primarily ifthey are large. Small diversions can easily be absorbed by existing spare capacity at other units; newconstruction or price increases would be unlikely. Large diversions could lead to capacity scarcity andincreased prices, changing the viability of the existing units.

Exhibit 4-20 presents estimated waste diversions under the low price pass-through scenario. Formost proposed MACT options, diversions are approximately four percent of the current combusted wasteuniverse with the bulk of diversions coming from the on-site incinerator sector. While these shifts couldhave noticeable localized effects on some combustion facilities, the impact on the market overall is likelyto be small. There is currently adequate spare capacity in all sectors to absorb these shifts.

Options 2a and 4, requiring that cement kilns install PIC controls, will lead to higher wastediversions as more kilns exit the market. Even under these options, however, diversions comprise onlyabout 10 percent of combusted wastes.

4.2.3.2 High Price Pass-Through Scenario

The high price pass-through scenario assumes that all of the new compliance costs for the lowest-cost combustion sector (again cement kilns) could be shifted to generators in the form of higher combustionprices. EPA examined increases in the weighted average price charged by each combustion unit, reflectingits particular mix of liquids and solids burned. The weighted average price increase assumed in the highpass-through scenario ranges from $66 to $87 per ton, depending on proposed MACT option. At theextreme, proposed MACT options 2a and 4 assume an increase of over $380 per ton. As shown above inExhibit 4-6, even with the large cost increases under PIC controls, cement kilns remain the lowest costsector. This is driven by a more than $700 per ton cost advantage in the baseline.45

EPA believes, however, that price increases on the order of $380 per ton would likely lead towidespread changes in current waste management behavior, especially in terms of waste minimization andnon-combustion alternatives. While the Agency does not think that such an increase would be sustainedin the market, the results for PIC controls even under the high price scenario are presented to demonstratethe degree to which greater price pass-through limits the impact of the rule on combustion facilities.Details on the price increases by sector are shown in Exhibit 4-21.

Because a higher price pass-through means that combustion units earn more on each ton of wastethey burn, they need to burn fewer tons in order to cover their costs. Thus, both short-term and long-termBEQs under the high price pass-through scenario are lower than in the low price pass-through scenario.As a result, market exit is expected to be lower.

EXHIBIT 4-20

QUANTITY OF HAZARDOUS WASTE THAT COULD BE DIVERTED FROM COMBUSTION FACILITIES IN THE SHORT TERM


Percentage of allBRS CombustedOn-siteCommercialCementHazardous WasteTOTALIncineratorsIncineratorsLWAKsKilns

3%88,16065,7904,8601,87015,640Original Floor

3%102,01079,6404,8601,87015,640Option 1b

4%108,27085,9004,8601,87015,640Option 1c

10%292,30058,8104,8601,870226,760Option 2a

5%139,77085,9004,86033,37015,640Option 2b

4%108,27085,9004,8601,87015,640Option 3

10%306,53058,8104,86016,100226,760Option 4

Notes:Compliance costs include CEM costs.1.Combusted hazardous waste reported to BRS in 19912.

2,977,355excluding tonnage burned in on-site boilers:Totals may not add due to rounding.3.

EXHIBIT 4-21

WEIGHTED AVERAGE COMBUSTION PRICE PER TON ANDINCREASE IN PRICES DUE TO ASSUMED PRICE PASS THROUGH


CommercialSector with

LowestOn-siteCommercialLWACementTotal cost/tonIncineratorsIncineratorsKilnsKilnsOptions

Current weightedcement kilns$580$646$188$178average price

Increase in price due to compliance costs passed through

cement kilns$66$66$66$66Original Floor


cement kilns$87$87$87$87Option 1c

cement kilns$381$381$381$381Option 2a




Notes:Compliance costs include CEM costs.1.Median compliance costs per ton exclude units currently2.not burning hazardous waste.The commercial sector with the lowest total cost per ton (baseline 3.+ compliance cost) drives the assumed increase in combustion prices.Prices for on-site incinerators reflect the cost per ton of off-site treatment that4.generators avoid by burning the waste on-site.Weighted average price per ton = (solids percentage of total waste burned in each5.sector x solids price) + (liquids percentage of total waste burned in each sector xliquids price).


4.2.3.2.1 Short-Term BEQ

EPA's analysis indicates few major changes in the percentage of facilities meeting their short-termBEQ between the low and high price pass-through scenarios. As shown in Exhibit 4-22, the number ofunits meeting BEQ under the high scenario in the cement kiln, LWAK, and on-site incinerator sectorsincreases by between five and 10 percentage points. Commercial incinerators show no difference; the unitsexpected to close burn small waste quantities and will close under a wide range of conditions. The largestchanges occur in the proposed MACT options that require PIC controls. Were cement kilns able to passthrough their entire compliance costs to generators, the percentage of cement kiln units meeting BEQwould jump from 49 percent in the low pass-through scenario to 73 percent.

4.2.3.2.2 Long-Term BEQ

As with the low price pass-through scenario, over the longer run, more units fail to meet BEQs.Shifts are largest in the incinerator sectors, again due to their higher baseline combustion capital (seeExhibit 4-23). About 10 percent of the commercial incinerators and 15 percent of on-site incinerators areexpected to stop burning wastes when their waste burning capital needs to be replaced.

Comparing the percentage of units meeting their long-term BEQ in the high and low scenariosshows that more units are viable in the high scenario. The increased viability, however, is not that great.The number of facilities meeting long-term BEQs under the high price pass-through scenario increases byonly 10 percentage points or so from the low price pass-through scenario. This once again illustrates theimportance of the quantity of wastes burned. Tonnage over which to spread fixed costs is a larger factorin meeting BEQ than is the price of combustion.

As with the short-term BEQ, proposed MACT options requiring PIC controls show larger changesbetween the low and high scenarios. The number of cement kilns and LWAK units meeting long-termBEQ under the high scenario increases by between 15 and 25 percentage points. Even incinerators arehelped, since they can match the price increases put forth by the kilns. In both incinerator sectors, thenumber of units meeting long-term BEQ under the proposed MACT options requiring PIC controls on BIFsincreased by more than 10 percentage points over the low price pass-through scenario. As stated above,however, it is unlikely that price increases of this magnitude would be sustained in the market.

4.2.3.2.3 Expected Cessation of Hazardous Waste Combustion at the Facility Level

Facility-level cessation of hazardous waste combustion is presented in Exhibits 4-24 and 4-25. Formost MACT options, only the commercial and on-site incinerator sectors show any differences betweenthe short-term and long-term behavior. With the exception of incinerator closures under proposed MACToptions 2A and 4 (requiring PIC controls), facility-exit decisions are unaffected by which above-the-flooroption is chosen.

EXHIBIT 4-22

PERCENTAGE OF COMBUSTION UNITS MEETING SHORT TERM BEQ AFTER CONSOLIDATION(Percentage of combustion units; includes units not burning waste in the baseline)




31%6%63%18%0%82%62%0%38%16%0%84%Original Floor

31%7%61%18%0%82%62%0%38%16%0%84%Option 1b

30%11%59%18%0%82%62%0%38%20%0%80%Option 1c

24%0%76%18%0%82%46%0%54%24%2%73%Option 2a

30%11%59%18%0%82%62%0%38%20%0%80%Option 2b

30%11%59%18%0%82%62%0%38%20%0%80%Option 3

24%0%76%18%0%82%46%0%54%24%2%73%Option 4




EXHIBIT 4-23

PERCENTAGE OF COMBUSTION UNITS MEETING LONG TERM BEQ AFTER CONSOLIDATION(Percentage of combustion units; includes units not burning waste in the baseline)




50%4%46%29%0%71%62%0%38%20%0%80%Original Floor

50%4%46%29%0%71%62%0%38%20%0%80%Option 1b

52%2%46%29%0%71%62%0%38%22%0%78%Option 1c

37%2%61%18%0%82%54%0%46%27%4%69%Option 2a

52%2%46%29%0%71%62%8%31%22%0%78%Option 2b

52%2%46%29%0%71%62%0%38%22%0%78%Option 3

37%2%61%18%0%82%54%0%46%27%4%69%Option 4




EXHIBIT 4-24

NUMBER OF COMBUSTION FACILITIES LIKELY TO STOP BURNING HAZARDOUS WASTE IN THE SHORT TERM(net of facilities not currently burning waste)





44412Original Floor

48412Option 1b

51412Option 1c

24413Option 2a

51412Option 2b

51412Option 3

24413Option 4

Notes:Compliance costs include CEM costs.1.Percentage of facilities currently not burning waste in the cement kiln, LWAK, and2.commercial incinerator sector; or burning less than 50 tons per year in the on-siteincinerator sectors. Some additional units may be nonviable in the baseline, leadingus to overestimate closures due to the proposed MACT.

EXHIBIT 4-25

NUMBER OF COMBUSTION FACILITIES LIKELY TO STOP BURNING HAZARDOUS WASTE IN THE LONG TERM

(net of facilities not currently burning waste)





72612Original Floor

72612Option 1b

72612Option 1c

48414Option 2a

72622Option 2b

72612Option 3

48414Option 4

Notes:Compliance costs include CEM costs.1.Number of facilities currently not burning waste in the cement kiln,2.LWAK, and commercial incinerator sectors; or burning less than 50tons per year in the on-site incinerator sector. Some additional unitsmay be nonviable in the baseline, leading us to overestimateclosures due to the proposed MACT.


Of facilities currently burning wastes, EPA estimates that two to three cement kilns, one LWAK,four commercial incinerators, and between 24 and 51 on-site incinerators will exit the hazardous wastecombustion market in the short-term. The greatest degree of exit under the high pass-through scenario isat facilities that burn very little waste now.

While higher prices do reduce facility exit somewhat, the change is not large. Cement kiln exitdeclines only when PIC controls are required (from nine to three). LWAK exit at the floor is unchanged,but drops by as many as three facilities for options requiring PIC controls. Price pass-through influencescommercial incinerator exit decisions only for the PIC-control options, and even there only in the long run.Only in the on-site incinerator sector do higher prices keep more facilities burning hazardous waste to anysignificant degree over the long-term, with about more on-site facilities staying in the market in the highscenario than in the low scenario. For options with PIC controls, as many as 24 additional facilities remainin operation, an increase of 50 percent.


Under the high price pass-through scenario (shown in Exhibit 4-26), the impact on operating profitsis substantially reduced from the 25 percent scenario. For cement kilns, absolute profit levels remainconstant as they recover full costs under all options. Improvements also occur in all other sectors. Boththe LWAK and commercial incinerator sectors see dollar operating profit margins rise because priceincreases at the level of compliance costs for the most efficient combustion sector (cement kilns) exceedthe compliance costs for the LWAKs and commercial incinerators under many MACT options. As withthe low price pass-through scenario, gains are largest for incinerators under options 2a and 4 where PICcontrols are required at the BIFs.

Under some MACT scenarios within the LWAK and on-site incinerator sectors, absolute operatingprofits decline even with 100 percent pass-through. The most extreme example of this is MACT option2b in the LWAK sector, where PIC controls are required for LWAKs but not for cement kilns. In thisscenario, price increases are small relative to the LWAKs' compliance cost per ton, resulting in a largedecline in operating profits. For the on-site incinerator sector, average absolute margins decline under mostoptions. This result is driven primarily by the number of combustion units that remain in the market.Higher price pass-through allows more marginal units to remain in business in the short-term, operatingat a level that does not recover invested capital.

While the situation in terms of absolute operating profits improves for most sectors, the impact onprofit margins continues to be negative. This is because the same dollar profits are being earned on aservice that sells for more money (the old price plus price pass-through). Essentially, the combustionoperator must invest more funds (to buy new emissions control equipment) in order to earn the same moneyas before the rule. Constant profits coupled with higher levels of investment may result in lower returnson assets for these firms, increasing the difficulty of raising capital somewhat.

EXHIBIT 4-26

CHANGE IN AVERAGE OPERATING PROFITS PER TON OF HAZARDOUS WASTE BURNED FROM THE PROPOSED MACT


On-site IncineratorsCommercial IncineratorsLWA KilnsCement Kilns% Margin afterOperating Profit Margin% Margin afterOperating Profit Margin% Margin afterOperating Profit Margin% Margin afterOperating Profit Margin

the Rule% Change$ Changethe Rule% Change$ Changethe Rule% Change$ Changethe Rule% Change$ ChangeOptions

24%-24%($23)25%15%$3767%-11%$2251%-27%$0Original Floor

25%-24%($24)25%14%$3667%-11%$2251%-27%$0Option 1b

24%-28%($29)24%8%$3060%-21%$753%-26%$0Option 1c

9%-248%$10238%76%$21725%-59%($6)31%-59%$0Option 2a

23%-31%($35)23%5%$2532%-60%($98)53%-26%$0Option 2b

24%-28%($29)24%8%$3052%-32%($21)53%-26%$0Option 3

10%-256%$10638%76%$21720%-67%($32)31%-59%$0Option 4

Notes:Compliance costs include CEM costs.1.Operating Profits = (weighted average price per ton + weighted average energy savings per ton + assumed price increase due to compliance costs passed through) - (average baseline costs2.per ton + average total annual compliance cost per ton). Assumed price pass-through is a set percentage (shown at the top of this exhibit) of the median compliance cost for the mostefficient combustion sector. As this is a static model, we have capped the price pass-through using the combustion units expected to remain burning hazardous waste even though theoriginal pass-through value included some units expected to stop burning. This is a better approximation of the impetus combustors have to raise prices, though it is not a precise predictor. To address uncertainty regarding the amount prices will rise, a variety of price increase scenarios were used. All other averages were calculated after consolidation, and include only thoseunits that continue to burn hazardous waste.Operating profits exclude overhead, other administrative costs, and taxes. Actual after-tax profits will be lower.3.Percentage Operating Profit Margin = average operating profits per ton / (weighted average price per ton + assumed price increase due to compliance costs passed through). Percentage profit margin after the rule is4.calculated using the same formula with post-rule operating profits and prices.Change in operating profits per ton = Post-rule operating profits per ton - baseline operating profits per ton. Percentage change in operating profits margin = (post-rule operating profits margin - baseline5.operating profits margin) / baseline operating profits margin. Baseline operating profit margins for units remaining open after consolidation can be calculated by dividing the percentage profit margin afterthe rule by one plus the percentage change in the operating profit margin. For consistency, baseline values have been calculated using the median compliance cost per ton for facilities that remain inoperation after the rule for each MACT option.

The large declines in percentage margins (such as under options 2a and 4 for the on-site incinerators)46

despite large increases in absolute operating profits are due to negative operating profits in the baseline,implying many burners are treating existing capital equipment as "sunk".

The price increase after consolidation is capped at the median compliance cost per ton for cement kilns47

that remain in the hazardous waste burning market. While this provides a better measure of the impetusto raise prices after the rule, it generates a price increase of about $90 per ton less than the mediancompliance cost for the entire universe of hazardous waste burning cement kilns before consolidation, andtherefore depresses the calculated operating margins. The impacts from this capping are largest underoptions 2a and 4.


Average margins in the on-site incinerator sector after the rule are relatively low. This outcomeis the result of a combination of two factors. First, there are many units operating at levels that do notrecover full capital costs. Second, the average is somewhat skewed downward by the capping of price46

increases based on the median cement kiln compliance cost for the cement kilns that remain in the wasteburning market after consolidation (see Exhibit 4-26, note 3). By capping the post-consolidation priceincreases at this cost, the calculated operating profit margins for the on-site incinerator sector decline undermost options. As stated throughout the report, EPA does not think 100 percent price pass-through under47

options 2a and 4 is likely to be sustained in the market.

4.2.3.2.5 Waste Diversions

As discussed above, while a large number of facilities decide to stop burning hazardous wastes,the waste quantities they currently manage are quite small (see Exhibit 4-27). EPA estimates that less thanthree percent of the wastes currently burned at combustion units regulated by the proposed MACTstandards will be diverted due to facility closure. Since the PIC control options assume a larger priceincrease is matched by the other combustion sectors (since they have higher combustion costs than thecement kilns), more facilities stay in the combustion business under PIC control than under other proposedMACT options; hence, projected waste diversions are lower. As stated above, we do not have a highdegree of confidence in this particular conclusion. With this exception, we expect minor waste diversionsunder our high price pass-through scenario across proposed MACT options.

4.2.3.3 Zero-Percent Price Pass-Through Scenario

To bound the impact of the proposed rule on combustors, EPA also evaluated a zero-percent pricepass-through scenario. This scenario assumes that new compliance costs from the rule must be paidentirely from existing profits; there will be no increase in the price of combustion services. The zero-percent price pass-through scenario models a market situation where there are many low-cost wasteminimization and non-combustion alternatives for waste streams that are currently combusted, preventingprice increases. As the price increases under the low price-through scenario were quite small, differencesbetween the that scenario and the zero-percent scenario are also relatively small. A more detailedcomparison between the low and zero scenarios can be found in Appendix G.

EXHIBIT 4-27

QUANTITY OF HAZARDOUS WASTE THAT COULD BE DIVERTED FROM COMBUSTION FACILITIES IN THE SHORT TERM


Percentage of allBRS CombustedOn-siteCommercialCementHazardous WasteTOTALIncineratorsIncineratorsLWAKsKilns

3%81,66059,2904,8601,87015,640Original Floor

3%86,85064,4804,8601,87015,640Option 1b

3%93,11070,7404,8601,87015,640Option 1c

2%58,87014,8904,8601,87037,250Option 2a

3%93,11070,7404,8601,87015,640Option 2b

3%93,11070,7404,8601,87015,640Option 3

2%58,87014,8904,8601,87037,250Option 4

Notes:Compliance costs include CEM costs.1.Combusted hazardous waste reported to BRS in 19912.

2,977,355excluding tonnage burned in on-site boilers:Totals may not add due to rounding.3.

Data on commercial BIFs and incinerators are based on surveys conducted by EI Digest, a trade journal.48

Data are generally one year old, and are sometimes reported at the facility rather than the combustion unitlevel. Data on on-site incinerators, and on the mix of liquids and solids burned, are based on EPA data suchas the BRS. Data are between two and four years old.

While average values also overstate revenues at some facilities (making certain units appear more49

profitable than they are in reality), few facilities appear so close to BEQ that they would fall below it iftheir revenues fell slightly.


4.2.3.4 Summary of BEQ Analysis

A number of key conclusions arise from our BEQ analysis. First, with the exception of PICcontrols, the proposed MACT option chosen has little impact on the number of facilities that stop burninghazardous waste; most impacts arise in achieving the proposed MACT floor. Second, the quantity ofwastes burned at a facility is the most important determinant of whether a combustion unit can meet itsBEQ. The ability to raise combustion prices does help facilities meet their BEQ in all sectors, but pricesare less important than tons burned. Third, additional market exit can be expected in both the commercialand on-site incinerator sectors as waste burning capital needs to be replaced. In contrast, cement kilns andLWAKs that meet their short-term BEQ are likely to continue burning hazardous wastes for the foreseeablefuture.

In terms of the number of facilities that stop burning hazardous wastes, the BEQ analysis illustratesimportant differences across sectors. Few cement kilns or commercial incinerators will stop burninghazardous wastes even under the zero price pass-through scenario. While some LWAKs and many on-siteincinerators are expected to stop combusting waste, most of these units burn very little waste now. As aresult, waste diversions as combustion facilities close are expected to be very small under both high andlow scenarios and under all regulatory options except PIC controls with low price pass-through, wherediversions reach up to ten percent of the current combustion universe.

In general, although this analysis is subject to numerous uncertainties, EPA believes it tends tooverstate closures in the commercial sector. The key uncertainties are as follow:

! As noted above, not all market exit can properly be attributed to the new proposedMACT standards. In addition to permitted facilities that currently burn nohazardous waste (and have been excluded from the Exhibits), there are somefacilities that burn extremely small tonnages, and, based on available data, appearto burn at a loss even under the current regulatory environment.

! BEQ calculations are sensitive to data on waste quantities burned, data that maynot be fully up-to-date. Units reported to have burned very low quantities of48

wastes in the past are likely to have either increased tonnages (if they had been instart-up mode) or exited the market already. Either change would reduce thelikelihood of market exit attributable to this rule.

! BEQ calculations are also sensitive to combustion prices. We rely on nationalaverages, and therefore may understate waste burning revenues at combustionunits that appear to be below BEQ .49


All of these factors make it more likely that this analysis overstates closures due to the rule. As a result,EPA believes that actual market exit in the commercial sector will be lower than our model predicts,although the model plant approach may over- or understate costs and the likelihood of exit for a specificcombustion unit.

One caveat is that declining profits on hazardous waste combustion services provided by cementkilns may reduce the ability of kilns to cross subsidize marginal cement operations with hazardous wasterevenues. EPA does not expect this to be a major issue because cement markets are extremely healthynow, and because operating profits on waste burning remain quite strong even under the low price pass-through scenario (although margins do fall).

In the on-site incinerator sector, uncertainties work in both directions. In terms of overstating thenumber of incinerators likely to stop burning hazardous wastes, three factors are in operation:

! Waste quantity burned data for on-site incinerators is four years old and is self-reported by combustors. Inaccuracies could be substantial.

! Operators of some on-site incinerators may continue to operate units at a loss toavoid liabilities associated with off-site shipments.

! Some on-site incinerators may spread the fixed costs of combustion over bothhazardous and non-hazardous wastes burned at the incinerator.

In contrast, one important factor suggests that our model may understate closures for on-site incinerators.In every model plant category, EPA assumes that the uncharacterized universe of facilities follows thesame pattern as the characterized universe. There are some indications that the on-site incinerators thatwere not characterized burned smaller waste quantities than those that were. If this is the case, the unitswould be more likely to fall short of their BEQ. Without better data, the Agency cannot determine the neteffect of these four factors.

4.2.4 Total Costs of the Proposed MACT Standards

Chapter 3 presented the pre-tax engineering costs of the rule, assuming that all facilities wouldcomply with the rule. The breakeven quantity analysis identifies how many combustion units are likelyto stop burning hazardous wastes, rather than comply with the rule. Because these units do not need toinvest in pollution controls, the total cost of the rule decreases. Wastes diverted as a result of market exitare assumed to be managed by the remaining units, which would probably increase costs slightly.

Note that for an on-site incinerator, the "revenue" impact of burning 25 percent less waste is the loss of50

implicit revenue when waste is sent to a commercial combustion facility.


Because the estimate of total pre-tax costs is affected by the number of combustion units that exitthe market, the degree to which compliance costs can be passed through to generators in the form of higherprices will affect the overall cost of the rule. EPA therefore prepared two estimates, one for the low pricepass-through scenario and one for the high price pass-through scenario. Exhibits 4-28 and 4-29 presentthese estimates. The Agency also evaluated a zero-percent pass-through scenario to bound the expectedcosts of the proposed rule. Total expected costs are approximately 20 to 30 percent below the engineeringcosts of the rule that were presented in Chapter 3. Because compliance costs are tax-deductible businessexpenses, the cost of the rule to firms will be a further 20 to 30 percent below this, depending on the firm'smarginal tax rate.

The impact of market exit on costs incurred is even clearer when viewed on a cost per ton basis.Exhibits 4-30 and 4-31 present average total annual compliance costs per ton under both the low and highprice pass-through scenarios. While there remain differences across sectors, they are relatively small. Themost noticeable difference between the after-consolidation numbers and those before consolidation (shownin Exhibit 4-7) is in the on-site incinerator sector. Costs per ton after consolidation are in some cases lowerthan for BIFs, where before consolidation they were in the thousands of dollars per ton. This shift is dueto the exit of the least efficient facilities (i.e., those spreading costs over little or no waste). Again, costsvary minimally across all MACT options with the exception of PIC controls. Also excluding PIC controls,the compliance costs per ton in the zero-percent price pass-through scenario are very similar to the lowscenario.

4.2.5 Waste Feed Reduction Compliance Option

As an alternative to installing new pollution control equipment to comply with the proposed MACTstandards, facilities may be able to comply by reducing waste feed. However, data suggest that the numberof facilities that will institute feed reductions is likely to be small. First, even if one assumes that feedreduction is costless for the combustion facility (which it is not), a relatively limited number of unitsactually shift to a less costly model plant when the 25 percent feed reduction is pursued. As shown inExhibit 4-32, the number of combustion units moving to a new model plant varies by regulatory option.Overall, it appears that on-site incinerators and cement kilns (under some regulatory options) have thegreatest likelihood of reducing pollution control expenditures with feed reductions. Virtually nocommercial incinerators will find it advantageous to reduce feed. In general, the maximum number ofcombustion units that change model plant assignment is 18 (under Option 4), roughly 14 percent of all theunits in the model plants analysis.

The incentive to pursue feed reductions will be further limited by the fact that such reductions willcurtail the revenue earned on waste burning. As noted earlier, combustors may choose to simply burn 25percent less waste, reducing revenues proportionately. In some cases, the revenue loss that follows from50

EXHIBIT 4-28

TOTAL ANNUAL PRE-TAX COMPLIANCE COSTS (millions)AFTER UNIT CONSOLIDATIONS


PercentageDifference from

EngineeringOn-siteCommercialLWACementCostsTotalIncineratorsIncineratorsKilnsKilnsOptions

-26%$140$70$17$3$51Original Floor

-27%$152$74$19$3$56Option 1b

-29%$175$84$22$4$65Option 1c

-49%$286$98$26$12$150Option 2a

-32%$191$93$26$7$65Option 2b

-29%$176$84$22$5$65Option 3

-49%$285$98$26$10$150Option 4

Notes:Compliance costs include CEM costs.1.Compliance costs after consolidation include the costs for those units that will continue to burn2.waste, as well as the shipping and disposal costs (after the assumed price increase) for on-siteincinerators that decide to stop burning wastes on-site. Other types of combustion units that stopburning wastes do not incur compliance costs and therefore are excluded.Because compliance costs are tax-deductible, the portion of pre-tax costs borne by the firm would3.be between 70 and 80 percent of the values shown above, depending on the specific firm's marginaltax bracket."Consolidation" allows for non-viable combustion units to consolidate waste flows with other units at the4.same site, or to exit the waste burning market. As a result, the number of combustion units incurringcompliance costs is reduced.

EXHIBIT 4-29

TOTAL ANNUAL PRE-TAX COMPLIANCE COSTS (millions)AFTER UNIT CONSOLIDATIONS


PercentageDifference from

EngineeringOn-siteCommercialLWACementCostsTotalIncineratorsIncineratorsKilnsKilnsOptions

-22%$147$73$17$3$55Original Floor

-23%$159$77$19$3$60Option 1b

-25%$185$89$22$4$71Option 1c

-28%$403$106$26$17$254Option 2a

-27%$206$97$26$12$71Option 2b

-25%$187$89$22$5$71Option 3

-28%$405$106$26$19$254Option 4

Notes:Compliance costs include CEM costs.1.Compliance costs after consolidation include the costs for those units that will continue to burn2.waste, as well as the shipping and disposal costs (after the assumed price increase) for on-siteincinerators that decide to stop burning wastes on-site. Other types of combustion units that stopburning wastes do not incur compliance costs and therefore are excluded.Because compliance costs are tax-deductible, the portion of pre-tax costs borne by the firm would3.be between 70 and 80 percent of the values shown above, depending on the specific firm's marginaltax bracket."Consolidation" allows for non-viable combustion units to consolidate waste flows with other units at the4.same site, or to exit the waste burning market. As a result, the number of combustion units incurringcompliance costs is reduced.

EXHIBIT 4-30

AVERAGE TOTAL ANNUAL PRE-TAX COMPLIANCE COSTS PER TON (Short Term - After Consolidation)



$67$29$44$58Original Floor

$63$30$44$62Option 1b

$67$36$51$70Option 1c

$111$41$164$199Option 2a

$74$41$175$70Option 2b

$67$36$72$70Option 3

$107$41$184$199Option 4

Notes:Compliance costs include CEM costs.1.Average compliance costs per ton exclude units 2.that are not likely to comply with the rule.Average on-site incinerator compliance costs include direct3.costs of meeting the new emission levels. Indirect costs tofacilities that stop burning wastes and must ship them off-sitefor management are not included.Because compliance costs are tax-deductible, the portion of4.pre-tax costs borne by the firm would be between 70 and 80percent of the values shown above, depending on the specificfirm's marginal tax bracket.

EXHIBIT 4-31

AVERAGE TOTAL ANNUAL PRE-TAX COMPLIANCE COSTS PER TON (Short Term - After Consolidation)



$89$29$44$63Original Floor

$90$30$44$69Option 1b

$95$36$59$79Option 1c

$156$41$264$279Option 2a

$102$41$164$79Option 2b

$95$36$87$79Option 3

$152$41$290$279Option 4

Notes:Compliance costs include CEM costs.1.Average compliance costs per ton exclude units 2.that are not likely to comply with the rule.Average on-site incinerator compliance costs include direct3.costs of meeting the new emission levels. Indirect costs tofacilities that stop burning wastes and must ship them off-sitefor management are not included.Because compliance costs are tax-deductible, the portion of4.pre-tax costs borne by the firm would be between 70 and 80percent of the values shown above, depending on the specificfirm's marginal tax bracket.

EXHIBIT 4-32

Number of Units Changing Model Plant Assignments After Waste Feed Reductions

TOTALPercent ofNumberOn-siteCommercialCementAll Unitsof UnitsIncineratorsIncineratorsLWAKsKilnsOptions

12%155037Floor

Above the Floor5%700341b

12%1591231c

12%1591232a

13%17111232b

12%1591233

14%18111334

Number of Units Included in the Model Plant Analysis:45Cement Kilns13LWAKs17Commercial Incinerators54On-site Incinerators

Personal communication with fuel blenders, including Brian Dawson, et al., 22 February, 1995. 51


the feed reduction will outweigh the savings on pollution control equipment; in other cases, the revenueloss is less than the savings. Exhibit 4-33 summarizes this effect. In the hypothetical example, the cementkiln would lose more waste burning revenue by feed reductions than it would cost to install the incrementalpollution controls. The kiln would therefore install controls. In contrast, the LWAK could save nearly$400,000 in compliance costs by reducing waste feeds. As shown in these two sample cases, dependingon facility-specific conditions, the combustor may or may not have the incentive to burn 25 percent lesswaste.

Exhibit 4-33

REVENUE LOSSES UNDER FEED REDUCTION COMPLIANCE OPTION

Example Regulatory Control if Burn if Burn 25%Facility Option 25% Less Waste Less Waste

Savings on Pollution Revenue Loss

Cement Kiln Floor $424,000 $894,000

LWAK 1c $611,000 $228,000

To achieve the feed reduction, commercial combustion facilities (but probably not on-site facilities)also have the option of requesting that blenders supply them with waste that contains 25 percent less metalsand chlorine. As noted, fuel blenders indicate that there would be a cost associated with supplying thislower-concentration waste. Blenders estimated a 50 percent increase in the cost of waste handling,assuming that they can meet demand by segregation and dilution. Without knowing the baseline costs51

of waste blending, however, we are unable to estimate the magnitude of the new costs that blenders wouldexperience. In addition, determining the economic effect of increased blending costs is difficult becausethe cost of supplying combustion facilities with lower-concentration waste would be split between threegroups:

! Combustion facilities could receive a lower price per ton for specially blendedwastes;

! Generators could bear a portion of the costs if blenders pass costs back to them inhigher prices; and

! Blenders may absorb a portion of the costs from profit.

Consequently, we cannot determine which of these outcomes will dominate. However, to the extent thatblenders attempt to pass costs on to combustors, combustors will have a reduced incentive to pursue thefeed reduction option.

Based on "Hazardous Waste Boilers and Industrial Furnaces (BIFs) with Either a Metals Reclamation52

Exemption or Small Quantity Burner Exemption," U.S. EPA Permits and State Programs Division,February 23, 1994. Metals reclamation facilities are not included in our analysis.


Overall, the feed reduction compliance option is likely to have a minimal effect on the total costof the rule. First, as noted in Chapter 3, the feed reduction option has very limited impacts on the overallcosts of compliance (roughly two to five percent). As shown here, this is because few facilities havesufficiently low emissions to install more modest pollution control when feed reductions are introduced.Finally, as we have demonstrated, the savings on pollution control equipment will, in some cases, beexceeded by the lost revenues; this further limits the likelihood that combustors will pursue the option.

4.2.6 Effects of Removing Small Quantity Burner Exemption

As noted earlier, the new proposed MACT rule will temporarily remove the BIF rule exemptionfor small quantity burners. Our evaluation of the impact of this action relies on a comparison of wastemanagement costs to overall facility sales for the affected facilities. Exhibit 4-34 reports the results of thisanalysis. The facilities listed are those that applied for the small quantity burner exemption under the BIFrule as of February, 1994. As shown, roughly half of the firms claiming the exemption are in SIC 491152

(electrical services). Furniture manufacturers are also prominent on the list. While employee data wereavailable for only a subset of facilities, the data suggest that firms claiming the exemption range from large(e.g., 2,400 employees) to very small (e.g., one employee), with an average of 350 employees.

Overall, it appears that removal of the exemption would have a small impact on all but a fewaffected facilities. As shown, the estimated lower bound cost of off-site management at a fuel blender isapproximately $24,000 per year while the upper bound cost is about $117,000 per year. Sales data areavailable for 23 of the 82 affected facilities. In the case of extremely small facilities, the costs could besubstantial; four facilities show upper bound costs greater than four percent of total revenues. However,for most facilities the off-site management costs that would result when the exemption is removed representa minor portion of facility revenues. On average, the costs are between 0.4 and 1.9 percent of facilityrevenues. The likelihood of adverse economic impact is especially limited for the electrical servicesfacilities, which make up much of the SQB universe. These facilities tend to be larger and will likely beable to absorb the relatively minor increases in waste management costs.

4.3 SUMMARY OF ECONOMIC IMPACTS ON AFFECTED GROUPS

This section summarizes the impacts of the proposed MACT standards on the combustion market,and then presents the impacts on each key market participant: cement kilns, commercial incinerators,lightweight aggregate kilns, on-site incinerators, hazardous waste generators, and fuel blenders. Althoughon-site boilers are not covered under the proposed rule, all other combustion sectors are. Therefore, EPAthought it important to summarize how the exempt boilers might interact with the other market sectorssubject to the rule. Finally, we provide context for evaluating the proposed MACT standards by examiningthe impact of a number of other pending market changes on competitive dynamics in the combustionindustry.

EXHIBIT 4-34COST IMPACT OF REMOVING SMALL QUANTITY BURNER EXEMPTION

91.2Estimated Waste Per Facility (tons) =$260Price Paid to Fuel Blenders for Liquids =

$1,280Price Paid to Fuel Blenders for Solids =$23,712Off-Site Management Cost (Lower Bound) =

$116,736Off-Site Management Cost (Upper Bound) =

Offsite Waste ManagementCost as a Percent of SalesEstimatedNumber ofSIC

UpperLowerSalesEmployeesCodeSTLocationFacility Name

FLCitrus CentralNCNCDHR - Caswell CenterNCVaughn-Bassett - Elkin

1521INMeromHoosier Energy20.2%4.1%$577,00041611WIEau ClaireEau Claire Asphalt

0.1%0.0%$198,683,0007002032NCMaxtonCampbell Soup Company0.0%0.0%$569,820,0002,4002033OHNapoleonCampbell Soup Company0.2%0.0%$51,000,0002702041INDecaturCentral Soya0.2%0.0%$70,443,0001012075OHBellevueCentral Soya

24NCMasonite2434NCIXL Furniture Co.

25NCCollingwood Furniture2511NCOld FortEthan Allen

1.1%0.2%$10,690,0001402511NCSpruce PineEthan Allen2511NCStanley Furn. - Lexington2511NCStanley Furn. - Robbinsville2512NCHickory Chair

0.1%0.0%$93,399,0007502531NYWaylandGunlocke0.9%0.2%$13,042,0001502599NCGregson Furniture

2599NCThomasville Furniture1.5%0.3%$8,000,000672631WVWallsburgBanner Fiberboard

2819SCDOE - Savannah River2865LALake CharlesOlin Chemical

5.8%1.2%$2,000,000202869PASomersetCarbose2869OHDoverUnion Camp

0.0%0.0%$352,362,0001,7003011INWoodburnUniroyal Goodrich Tire Company3221FLAnchor Glass Container3221GAAnchor Glass Container

71.2%14.5%$164,00013412NYPort CraneTri-Cities Barrel Company3489MNMinneapolisFMC Corp Naval Systems Division3999TNBruce Hardwood Floors4911PAAdrianAllegheny Power - Armstrong Station4911PAMasontownAllegheny Power - Hatfield Station4911AZChallaArizona Public Services4911NCCP&L - Cape Fear4911NCCP&L - Lee4911NCCP&L - Mayo

1.9%0.4%$6,084,000154911NCCP&L - Roxboro4911NCCP&L - Sutton4911NCCP&L - Weatherspoon4911WACentraliaCentralia Steam Electric Plant4911ILGrundy CountyCommonwealth Edison Collins Street4911NDUnderwoodCooperative Power Association4911MNFox LakeInterstate Power Fox Lake Station4911IASioux CityIowa Public Service4911KSLacgyneKansas City Power & Light

0.2%0.0%$67,330,0001664911MOClintonKansas City Power & Light

EXHIBIT 4-34COST IMPACT OF REMOVING SMALL QUANTITY BURNER EXEMPTION

91.2Estimated Waste Per Facility (tons) =$260Price Paid to Fuel Blenders for Liquids =

$1,280Price Paid to Fuel Blenders for Solids =$23,712Off-Site Management Cost (Lower Bound) =

$116,736Off-Site Management Cost (Upper Bound) =

Offsite Waste ManagementCost as a Percent of SalesEstimatedNumber ofSIC

UpperLowerSalesEmployeesCodeSTLocationFacility Name

4911MOWestonKansas City Power & Light0.0%0.0%$857,450,0004004911MOKansas CityKansas City Power & Light

4911WVMaidsvilleMonongahela Power - Ft. Martin Station4911WVWillow IslandMonongahela Power - Pleasants Station4911WVHaywoodMonongahela Powr - Harrison Station4911OKHarrahOG&E - Horseshoe4911OKMuskogeeOG&E - Muskogee4911OKOklahoma CityOG&E - Mustang4911OKKonawaOG&E - Seminole4911OKRed RockOG&E - Sooner4911NEOmahaOPPD North Omaha Station4911NEOmahaOPPD Unit One Station4911WYGlenrockPacific Power4911WYPoint of RocksPacific Power4911WYGillettePacific Power/Utah Power4911ORBoardmanPortland General Electric4911IADavenportRiverside Generating Station4911TXLubbockSW Public Service Center - Jones Station4911TXAmarilloSW Public Service Center - Nichola4911TXEarthSW Public Service Center - Plant X4911ILSpringfieldSpringfield CWLP

0.4%0.1%$30,420,000754911TXFort WorthTX - NM Power Company0.4%0.1%$32,042,000794911NDStantonUnited Power Association0.1%0.0%$106,267,0002624941IAMuscatineMuscatine Power & Water0.1%0.0%$106,267,0002624941IAMuscatineMuscatine Power & Water0.2%0.0%$74,936,0002005085CAHaywardAnchor Glass Container0.1%0.0%$206,364,0002205169NCOccidental Chemical

7349UTProvoGeneva Steel7537WVChapmanvilleTristate Auto Transmission

51.0%10.4%$229,00037538MNCambridgeSwede's Repair8011NEOmahaUniversity of Nebraska Medical Center

3.1%0.6%$3,768,000708221GAUniversity of Georgia9661FLKennedy Space CenterUnk.MDWilliamsportPEPCO - Paul Smith Station

1.9%0.4%Average =350Average =71.2%14.5%Maximum =

Sources:

1. PSPD, Hazardous Waste Boilers and Industrial Furnaces (BIFs) with Either a Metals Reclamation Exemption or Small QuantityBurner Exemption, February 23, 1994.2. Dun & Bradstreet3. EI Digest, September, 1994.4. IEc Analysis


4.3.1 Key Conclusions Applicable to Entire Combustion Market

EPA's proposed MACT standards will increase the cost of hazardous waste combustion across theindustry, and will lead some facilities to stop burning hazardous wastes altogether. The market impacts,however, are expected to be small. Only a small number of facilities are expected to stop burning wastesin all commercial sectors. While more on-site incinerators will cease waste burning, they tend to burn verylow quantities of hazardous waste now. Waste diversions from combustion are likely to be low as a result.The rule is expected to cost between $140 and $286 million per year. Costs jump to a high of $405 millionper year if PIC controls at kilns are required. New compliance costs per ton of waste burned are relativelysimilar across all commercial combustion sectors, suggesting that the rule will not trigger large changesin the competitive balance between sectors.

The quantities of hazardous waste currently combusted are the primary determinant of whethercombustion facilities will stop burning. Much of the market exit in all four market sectors is driven by lowquantities of waste over which to spread the fixed costs of compliance with the new standards. Since thefacilities that stop burning hazardous waste tend to be those currently burning low quantities, EPAestimates that only about three percent of the tonnage currently combusted will be diverted to new outlets.

The compliance cost per ton of waste burned varies little across proposed MACT options. As aresult, most of the costs of the rule, and market exit from the rule, occur in reaching the proposed MACTfloor. With the exception of options 2a, 2b (LWAKs only), and 4 that require PIC controls, the above-the-floor options do not noticeably change the impacts of the rule. Under proposed MACT options requiringPIC controls, compliance costs per ton rise dramatically for cement kilns and LWAKs, suggesting thatlarger changes in the combustion marketplace would occur.

Uncertainty with respect to the cost and availability of waste minimization and non-combustionwaste management alternatives for generators makes it difficult to predict how high combustion pricescould rise before generators switched substantial waste quantities to non-combustion venues. AlthoughEPA bounds this uncertainty by including a zero-percent price pass-through scenario, the impact of pricechanges on market exit is not large in most sectors. Price pass-through has a more direct effect on theprofitability of the remaining units.

Since firms that close combustion units in the face of the new rule do not incur complianceexpenditures for those units, the estimate of the total costs of the rule excludes the units likely to stopburning hazardous wastes. The total cost estimate includes compliance costs for all combustion units thatcontinue to burn hazardous wastes, plus any cost increases from having to manage wastes from closedfacilities in a more expensive manner. For all non-PIC proposed MACT options, total costs of the rule are22 percent lower than total engineering costs that assume all units comply with the rule. The costs for PICoptions decline by almost 50 percent once market exit is incorporated.

Kilns burn primarily high-Btu liquid solvent wastes. Solvent recycling would be a cost-effective53

alternative, limiting the degree to which kilns could increase prices. However, solid hazardous wastes aresuspended in the solvents prior to burning. Solvent recycling would eliminate this disposal outlet for solids,suggesting that prices on liquids with suspended solids could rise much more than prices on clean solvents.


4.3.2 Differential Impacts of the Rule by Market Sector

Although the rule will not cause large changes in sectoral competition, not all market participantswill be affected in the same way. This section summarizes the key impacts on the four combustionsegments, as well as on generators and fuel blenders. A comparison of important measures of impactacross sectors is presented in Exhibit 4-35.

4.3.2.1 Cement Kilns

Cement kilns currently have a very large baseline cost advantage over commercial incinerators dueto their ability to use large capital equipment (such as the kiln) for joint cement production/hazardous wastedestruction. While the proposed MACT standards will greatly change the baseline cost of burning at acement kiln, the existing cost advantage is so large that cement kilns remain the lowest-cost combustorsunder all proposed MACT scenarios.

Kilns likely to exit the market are those that tend to burn little waste now. Under proposed MACToption 1b, approximately 2 facilities currently burning hazardous wastes would stop doing so. Thisincreases to three facilities if prices could not rise at all. If kilns were required to install PIC controls (asunder proposed MACT options 2a and 4), and were not successful in passing the full costs of these controlsto generators in the form of higher prices, ten kilns are expected to stop burning hazardous wastes (17 ifprices couldn't rise at all). Although the kilns would have the lowest total combustion costs per ton evenwith PIC controls, the Agency believes that a tripling of combustion prices at the kiln would causegenerators to utilize alternative management methods.53

Despite the low expected market exit, kilns remaining in the market would experience a substantialdecline (on the order of 25 to 30 percentage points) in the profitability of hazardous waste combustion.This decline will make it more difficult for kiln operators to cross-subsidize cement production withhazardous waste profits. While strong cement markets now protect many of these marginal kilns, reducedhazardous waste profits could accelerate plant closures during the next industry downturn. EPA does nothave adequate information on marginal plants to assess the likelihood of such closures. Despite thedeclines in operating profits, BIFs continue to enjoy margins more than double those of commercialincinerators.

The market behavior of waste-burning cement kilns in the short-run and the long-run are is quitesimilar. Because the kilns have little in the way of baseline waste burning capital equipment, the long-termBEQ is quite similar to the short-term BEQ.


Exhibit 4-35

SECTORAL COMPARISON OF IMPACTS OF PROPOSED MACT OPTION 1B

Cement Commercial Lightweight On-SiteKilns Incinerators Aggregate Incinerators

Kilns

Total Compliance Costs ($Millions)- Before consolidation $71 $24 $9 $103- After consolidation, zero pass-through

scenario $51 $19 $3 $72- After consolidation, low pass-through

scenario $56 $19 $3 $74- After consolidation, high pass-through

scenario* $60 $19 $3 $77

Average Compliance Costs Per Ton ofHazardous Waste Burned- Before consolidation $90 $103 $139 $4,970- After consolidation, zero pass-through

scenario $59 $30 $44 $64- After consolidation, low pass-through

scenario $62 $30 $44 $63- After consolidation, high pass-through

scenario* $69 $30 $44 $90

Number of Combustion Facilities Likelyto Stop Burning Hazardous Wastes in theLong-Term**- Zero Pass-Through Scenario 3 6 1 89- Low Pass-Through Scenario 3 6 1 82- High Pass-Through Scenario 2 6 1 72

Notes:

* Costs rise in the high pass-through scenario because some units with higher compliance costsremain viable.

** Excludes facilities not burning waste in the baseline. Exit over long-term capital cycle;immediate market exit would be lower. Due to the large number of very small on-site burners,on-site incinerators burning less than 50 tons of hazardous waste per year excluded fromcalculation. For all other sectors, only units burning zero tons per year were excluded.

This market erosion may not necessarily occur in a linear fashion. For example, some kilns may not54

refocus on hazardous waste combustion expansion until the demand for cement slows.


4.3.2.2 Commercial Incinerators

Commercial incinerators face slightly lower compliance costs per ton burned than do cement kilns.As a result, commercial incinerators that remain in operation will become slightly more competitive withthe kilns. However, the dollar value of this benefit is about $40 per ton under all proposed MACT optionsnot requiring PIC controls for kilns, and is therefore unlikely to affect market competition significantly.We estimate that commercial incinerators continue to face average costs per ton of waste burned more than$700 higher than BIFs. Though their ability to handle waste streams that cannot currently be burned inkilns provides the incinerators with a protected niche, the longer-term erosion the commercial incinerator'smarket is likely to continue.54

Facility closure in the commercial incinerator sector is driven almost entirely by waste quantitiesburned. Four of the permitted facilities currently burning wastes are expected to close under both the highand the low price pass-through scenarios and are probably marginal even in the current market. Long-termclosures increase by two facilities in both scenarios over short-term exit. This increase is driven by thelarge baseline waste burning capital that must be recovered over the capital replacement cycle in order forfacilities to stay in the market.

Commercial incinerators also experience a decline in operating profit margins — of between 5 and20 percent in the low price pass-through scenario. As the declines are smaller than for cement kilns, therule will improve the competitive position of commercial incinerators relative to kilns. EPA does notanticipate that the changes are large enough to cause substantial changes in competitive dynamics.

4.3.2.3 Lightweight Aggregate Kilns

LWAKs face average compliance costs per ton of waste burned (before consolidation) are up totwice as high as those for cement kilns and commercial incinerators. This differential is overstated,however because over 30 percent of the permitted LWAKs do not actively burn wastes in their kilns. Oncenon-viable units exit the market, average compliance costs per ton drop below costs to cement kilns formost MACT options. Furthermore, EPA estimates that LWAKs have a $600 per ton baseline costadvantage over commercial incinerators. Before and after the rule, LWAKs retain a strong competitiveposition.

Kilns burning little or no waste now will not continue doing so in the face of the proposed MACTstandards. The substantial required investments in pollution control capital cannot be spread over the smalltonnage at these facilities, and the units will most likely stop burning hazardous wastes. Two LWAKs arelikely to exit the market, only one of which is actively burning wastes. As with cement kilns, exit increasesunder proposed MACT options requiring PIC controls unless these costs can be passed on to generators inthe form of higher prices. However, LWAKs currently burning reasonable quantities of hazardous wastewill remain strong competitors in the hazardous waste combustion market for both the short- and the long-run. While operating profit margins for remaining facilities do decline, dollar margins actually increaseunder many MACT options in the high scenario. Margins after the rule return to as much as twice thosein the incinerator sectors.


4.3.2.4 On-site Incinerators

On-site incinerators fall into two categories. At one end of the spectrum are facilities with highcapacity utilizations, burning large quantities of waste. These facilities are able to spread fixed investmentsso that costs per ton of waste combusted are reasonable. At the other end of the spectrum are a largenumber of incinerators that burn extremely low tonnages; it is these facilities that drive up the extremelyhigh pre-consolidation compliance costs per estimated ton of waste burned for this sector. To the extentthat these units also burn large quantities of non-hazardous wastes, some may remain viable under theproposed MACT standards. More likely, however, is that the units will stop burning hazardous waste evenif non-hazardous waste continues to be burned. This is because the BEQ on compliance costs alone oftenexceeds the tons of hazardous waste currently burned at the units.

Market exit in this sector is expected to be quite large — between 44 and 61 facilities in the short-term (assuming low price pass-through). As the incinerator capital itself requires replacement, over one-half of the on-site incinerators actively burning are expected to stop burning hazardous wastes. Sincequantities burned at each unit are quite low, the impact on these shutdowns on the combustion market isexpected to be minimal. Units that remain in the market, however, have operating profit margins similarto, or better than, commercial incinerators under most MACT options.

Decisions to stop burning hazardous waste at on-site facilities could result in increased investmentin waste minimization. Facilities that find it too expensive to come into compliance with the new MACTstandards will be faced with the choice of either sending their waste to a commercial combustion facilityor finding a non-combustion alternative. The waste minimization analysis conducted for the RIA suggeststhat many of these on-site generators may have waste minimization options that are less expensive thancommercial combustion. In these cases, the profit maximizing generator should choose waste minimizationover continued combustion.

4.3.2.5 On-site Boilers

On-site boilers are not covered under the proposed MACT standards. To the extent that the samewastes can be combusted in both on-site boilers and on-site incinerators, the rule makes boilers more cost-effective because they do not have to invest in new controls. Therefore, on-site boilers could provide alower-cost outlet for some of the wastes diverted from on-site incinerator closures (thereby reducing thecost of the rule). Alternatively, the existence of a boiler option could accelerate the closure of incineratorsas firms divert wastes even from incinerators that appear capable of meeting their BEQ. EPA has notestimated the magnitude of these impacts.

4.3.2.6 Hazardous Waste Generators

Hazardous waste generators are likely to see price increases for combusted waste streams. Withall combustion sectors facing increased costs, and with the compliance costs per ton similar acrosscommercial sectors, generators will have little flexibility to seek out a lower cost combustion sector.

EPA has proposed exempting the combustion of clean hazardous waste fuels from regulations governing55

hazardous waste combustion in general on the grounds that there is little difference between them andconventional fuels such as coal and oil.

Blending operations at commercial incinerators are often done on-site by incinerator employees, whereas56

blending at cement kilns is often run by outside firms.

In fact, blenders have stated that segregating waste streams to provide certain combustion units with57

lower metal wastes would be expensive. Personal communication with fuel blenders, op. cit., February22, 1995.


The size of the likely price increases is difficult to determine, and will differ by the type of waste.EPA anticipates that generators of clean solvents and lean waters will face lower price increases due to theavailability of non-combustion alternatives. High Btu-liquids without suspended solids can be reclaimed,or possibly burned in exempt facilities, both of which would constrain price increases that could be passedon to generators of the liquids. Land-ban solids and sludges could face more substantial increases.55

EPA does not have sufficient information on specific generating sectors to evaluate the impact ofprice increases on generator processes more specifically. However, the Agency does expect that a numberof waste minimization and non-combustion alternatives are available in the long-term should combustionprices rise significantly. For this reason, EPA price increases of between $20 and $90 per ton seem morelikely than larger increases. These changes would increase waste management costs for generators,although EPA has not analyzed the impact this could have on the prices of products produced by thesegenerators.

4.3.2.7 Fuel Blenders

Fuel blenders serve as intermediaries between generators and combustors in both the commercialBIF and commercial incinerator sectors. To the extent that the proposed MACT changes demand patterns56

of any one of these parties, blenders would need to react.

This analysis indicates that the impacts of the rule on blenders are likely to be relatively small.Very few combustion facilities can avoid capital equipment purchases by reducing the metals or chlorinein the hazardous waste fuels they burn, even if the new formulations are assumed to cost the same as theold. As a result, blenders are unlikely to be called on to change their fuel blends by more than a handful57

of customers. Waste diversions from closed facilities are also expected to be quite small. While somelocalized blender operations may need to ship wastes to different outlets, the magnitude of these changesnationally is unlikely to be large.

One exception involves proposed MACT options requiring PIC controls at commercial BIFs. Ifthe cost of compliance makes many kilns uncompetitive (a distinct possibility), blenders will lose their keywaste outlet. Alternatively, PIC controls could bifurcate the BIF market, with liquids exempted by cleanfuels regulations burned in non-waste burning BIFs (or recycled), while solids are burned in commercialincinerators. This could drastically reduce the demand for fuel blending services.

Much of the cement kiln dust can be recycled back into the cement. To avoid the buildup of alkalis in58

the product mix, however, some dust is vented from the process and disposed of. The CKD that is removedfrom the process is known as "net CKD."


4.3.3 MACT in Perspective

While the proposed MACT standard is the largest factor in terms of new costs to be borne by wasteburners, there are other factors that could have greater effects on the competitiveness of particular industrysectors. These factors are described below.

4.3.3.1 Cement Kiln MACT

The pending cement kiln MACT governs emissions from cement kilns even if they do not burnhazardous wastes. Although final MACT emission levels have not yet been set, the standard is likely torequire kilns to install emission controls for a variety of hazardous air pollutants. The more similar theMACT standards for non-burning kilns are to the standards for kilns that do burn hazardous wastes, themore air pollution control equipment that will become "standard" on any cement plant. Thus, theincremental costs of burning hazardous wastes in a kiln will be lower than are reflected in our cost models,suggesting that more kilns will continue to burn wastes. Kilns would become more competitive relativeto commercial incinerators.

4.3.3.2 Cement Kiln Dust Regulation

Pending changes to management requirements for cement kiln dust (CKD) will increase the costof managing residuals from cement production. This change will only affect combustion markets to theextent that net CKD generation is higher for waste burning kilns than for non-waste burning kilns becauseonly in this situation would CKD disposal be an incremental cost of waste burning. While EPA's Cement58

Kiln Dust report to Congress suggested that net CKD generation is significantly higher at kilns burninghazardous wastes, industry has argued that net CKD generation is unaffected by waste burning.

4.3.3.3 On-Site Boiler Exemption

The proposed MACT standards add a minimum of $65 per ton of waste burned in new compliancecosts that on-site boilers will not incur. While this cost advantage is unlikely to be large enough to threatenwaste flows now going to the commercial sector, on-site boilers could provide an attractive alternative toon-site incineration for certain waste streams. To the extent that significant waste flows move from on-siteincinerators to on-site boilers, closures of the incinerators could be higher than projected here and HAPemissions from boilers would rise.

"Firing Up Discussions on a Clean Fuels Exemption," EI Digest, April 1995, p. 1.59


4.3.3.4 Comparable Fuel Exclusion

EPA is proposing that certain hazardous waste fuels meeting specified levels for concentrationsof toxic constituents and physical properties that affect burning be excluded from the definition of solid andhazardous waste materials under the Resource Conservation and Recovery Act (RCRA). Generators thatcomply with sampling and analysis, notification and certification, and recordkeeping requirements wouldbe eligible to claim the exclusion. While these fuels are derived from hazardous wastes, they are "barelydistinguishable in constituents and burning properties from conventional fossil fuels." The exemption59

would allow some BIFs to continue burning hazardous fuels without complying with the new MACTstandards. While the spectrum of wastes that would be exempted has not yet been determined, it is likelythat a comparable fuel exclusion would affect the type and percentage of hazardous solids that could besuspended in liquid organic fuels by fuel blenders. The market for waste fuels could be split into one forextremely clean fuels and one for blended fuels containing higher levels of contaminants. The Agency doesnot have enough information at this time to evaluate the likelihood or implications of such a split oncombustion markets, but invites comment on this issue.

4.3.4 Summary

The proposed MACT standards should not change the competitive dynamics of hazardous wastecombustion markets, except under MACT options requiring PIC controls. Compliance costs per ton ofwaste burned are similar enough for facilities remaining in the market across all commercial outlets topreserve the existing competitive balance. While the rule will lead a significant number of combustionunits to exit hazardous waste combustion markets, these impacts are not skewed across combustion sectors,and are driven primarily by the low quantities of waste the units currently burn. Because facilities withlimited quantities are those that are least economically viable, facilities that stop burning hazardous wasteswill not cause large waste shifts.

While the exact price increases resulting from the rule cannot be predicted, the brunt of these priceincreases is likely to be borne by generators of solids and sludges, rather than of hazardous liquids. Undermost scenarios, price increases will be small to moderate. Under scenarios with large price increases,generators are likely to shift to non-combustion waste management alternatives for organic liquids and leanwaters. Fewer alternatives exist for solids and sludges.


COST-EFFECTIVENESS AND BENEFITS ANALYSIS CHAPTER 5____________________________________________________________________________________

A complete analysis of the effects of the MACT standards for combustion facilities must comparethe costs of the rule to its benefits. This chapter considers several different benefit/cost measures:

! Cost-Effectiveness: Provides estimates of the expenditures per unit reduction ofemissions for each HAP.

! Health Risk Benefits: Discusses the potential health benefits of this rule.

! Ecological Risk Benefits: Reviews the results of comparing surface waterpollutant concentrations and aquatic toxicity criteria.

! Economic Benefits: Provides a description of economic benefits potentiallyassociated with reduced combustor emissions, focusing on property value effects.

5.1 COST-EFFECTIVENESS

5.1.1 Cost-Effectiveness Methodology

EPA has developed a cost-effectiveness measure that examines cost per unit reduction of emissionsof each HAP. The two analytic components of this measure are:

! estimates of expenditures per HAP for each regulatory option; and

! estimates of emissions reductions under each regulatory option.

The discussion below describes the method for each of these components.

( 4040%60

x CI Cost) % ( 6040%60

x CI Cost)


5.1.1.1 Expenditures Per HAP

The "numerator" in the calculation of cost-effectiveness is the compliance expenditures associatedwith the specific HAP. Estimation of costs per HAP is complex because of the number of HAPs coveredin the rule and because the pollution control devices assumed in the model plants analysis frequently controlmore than one pollutant. Therefore, precise estimation of expenditures per HAP is not feasible. EPA hasdeveloped a simplified method that distributes costs to each HAP based on the following assumptions:

! For each HAP at each combustion unit, we calculate the percentage reduction inemissions required to reach the MACT standard and average the reductions bymodel plant group. For example, under MACT option 1b, a unit with dioxinemissions of 0.3 TEQ would need a 33 percent reduction to achieve the dioxinlimit of 0.2 TEQ.

! Control technologies are then assigned to each HAP. For example, carboninjection can control both mercury and dioxins. For each model plant, we attributethe cost of each technology to the specific HAPs that technology controls weightedby the average percent reduction for those HAPs required by the model plant. Forexample, if a model plant controls dioxin and mercury with a carbon injection (CI)system, the calculation would be as follows:

Percent dioxin reduction required to meet standard: 40 percentPercent mercury reduction required to meet standard: 60 percent

This splits the costs of the carbon injection system between dioxin and mercury;the same approach is used to develop a complete cost breakdown by HAP for eachtechnology within a model plant.

! For each HAP, the control costs for all technologies are summed within the modelplant, yielding a cost breakdown by HAP for each model plant. We then multiplythe cost breakdown by HAP for each model plant by the number of units in themodel plant; this yields a total cost by HAP of all units within a model plant group.

! We then sum the total cost by HAP of each model plant group for all model plantswithin each source category.

The scaling method used is the same as that described in Chapter 3.1


! Finally, we scale from the set of units in the model plants analysis up to the totalnumber of units in each source category to arrive at total expenditures per HAP.1

The total cost by HAP is then summed for each source category to obtain a totalcost by HAP for each MACT option.

5.1.1.2 Emissions Reductions

The "denominator" in the calculation of cost-effectiveness is the total mass emission reductionachieved when combustion facilities comply with the standards for the given regulatory option. Estimatingtotal emissions under each regulatory option requires two key assumptions:

! Combustion units that are already emitting below the standard for a given HAP donot change emissions of that HAP; and

! Combustion units with emissions exceeding the standard will control to thestandard.

The emission reductions are then calculated as the difference between the baseline total emissions reportedin Chapter 2 and the emissions under the regulatory option.

5.1.1.3 Caveats

Our method for calculating cost-effectiveness makes several simplifying assumptions. Mostimportantly, the method assumes that all facilities install control equipment and continue operating. Asdiscussed, a number of other responses to the MACT standards are possible. For example, some facilitiesmay cease waste burning in the face of increased compliance costs. However, it is difficult to trace theoverall effect that these reactions would have on either expenditures per HAP or on total emissions of eachHAP.

Beyond this broad caveat, other factors lead us to overstate expenditures per HAP. The assumptionthat units control emissions to the level of the standard likely leads us to overstate emissions becausefacilities employing emissions control equipment will likely achieve emissions concentrations below thestandard rather than exactly at the standard.

A final caveat concerns the method by which we apportion pollution control device costs by HAP.Costs are currently apportioned according to the percentage reduction required to meet the standard foreach HAP controlled by the device. In actuality, there is no "correct" approach for distributing costs.While the approach chosen is reasonable, it does not take into account engineering/technological issuesregarding the relative ease with which a device can control one pollutant versus another.


5.1.2 Cost-Effectiveness Results

Before discussing expenditures per unit reduction of each HAP, it is useful to first reviewinformation on total expenditures by HAP. As shown in Exhibit 5-1, the pollution control expenditures foreach HAP vary by pollutant and by regulatory option. The following findings appear most significant:

! For most regulatory options, control of dioxin, mercury, and semi-volatile metalscontributes the most to total costs. In options requiring improved combustion (PICcontrol) at cement kilns, expenditures on CO and total hydrocarbons (surrogatePIC measures) increase greatly and dominate total control costs.

! Expenditures on dioxin/furan control vary little by regulatory option, rangingbetween $37 and $56 million per year. This increase occurs between the floor andOption 1b, when a more stringent dioxin standard is introduced.

! Expenditures on mercury control range between about $28 and $85 million, withthe greatest increase coming between Option 1b and 1c, when protective mercuryrequirements are introduced.

As part of the cost-effectiveness measure, EPA also developed estimates of reduced pollutantloadings under each regulatory option. Estimates of these reduced emissions are presented in Exhibit 5-2.As shown, on a tonnage basis, CO, chlorine, HCl, PM and total hydrocarbons are the most significant.Dioxin, mercury, and metals are reduced in smaller quantities, but have much more significance from arisk reduction standpoint. In particular, dioxin emissions in the baseline are small -- about 2.3 pounds peryear. However, the highly carcinogenic nature of dioxin means that even reductions of two pounds per year(as experienced under the various regulatory options) may be significant.

Examination of the data on expenditures per ton reduced confirms that the more toxic pollutantsare often much more expensive to control on a per-ton basis. Exhibit 5-3 combines the cost per HAP andloadings reduction information to obtain the cost per unit of HAP reduced. The top portion of the exhibitshows the cost per unit of HAP reduced across all facilities while the bottom three sections show the resultsfor each facility type. Note that there is sometimes minor variation between options in cost per unit ofHAP reduced, even when the standards for particular pollutants do not change. This minor variation occursbecause the standard for a related pollutant (i.e., one controlled by the same device) may change, affectinghow the cost of the device is distributed to multiple pollutants (see methodology discussion above).

EXHIBIT 5-2

REDUCTION IN HAP EMISSIONS UNDER EACH OPTION (TONS)

ATF - 4ATF - 3ATF - 2bATF - 2aATF - 1cATF - 1bOriginal Floor

79,99612,80713,20279,99612,80712,80712,807CO954954950950950950950ChlorineNANANANANANANATotal Chlorine

3,8033,8031,2011,2011,2011,2011,201HCl27272727272727LVM9911111198Mercury

3,5883,5882,4972,4972,4972,4972,497Particulate80808080808080SVM

4,1772,5972,6544,1772,5972,5972,597THC0.0010.0010.0010.0010.0010.0010.001D/F

100,03131,26228,01996,33627,56727,56527,564TOTAL

Across all HAPs, cost-effectiveness is greatest in Option 4 because it involves greatly increased control2

of the pollutants for which cost per ton is lowest, e.g., carbon monoxide.


Unfortunately, the diverse nature of the HAPs precludes direct comparison of costs per HAPnormalized for toxicity. For example, dioxin is a carcinogen while mercury is a non-carcinogen. Likewise,the low- and semi-volatile metals categories are made up of different individual metals. However, thegeneral level of concern over dioxin and mercury exposure is consistent with the expenditures per tonreduced. Considering all facilities as a group, the following patterns are evident:

! Costs per unit of dioxin reduced are extremely high ($40,000 to $60,000 per gram)because of the minuscule amounts of dioxin released to the environment.

! Similarly, expenditures per ton of mercury reduced are between $3.7 and $9million.

! Expenditures per ton of metals reduced range roughly between $200,000 and$300,000.

For other pollutants, expenditures per ton reduced are lower. This is primarily a function of thelarge loadings reduction that is experienced. For example, despite large overall expenditures on carbonmonoxide control in Options 2a and 4, the equally large tonnage reductions result in a limited cost per tonreduced (about $2,000).

As shown in Exhibit 5-3, the pattern in costs per unit of HAP reduced varies somewhat when welook at different facility types. For example, at all facility types, dioxin and mercury control are costly;however, they are more costly on a unit basis at incinerators than at cement kilns. Conversely, control ofsemi- and low-volatile metals is more costly at cement kilns than at incinerators.

Another noteworthy pattern in the data is the change in cost per ton reduced when more stringentstandards are introduced. For example, the cost per ton of dioxin control rises when we move from thefloor to the above-the-floor options (moving from a 0.5 ng/dscm standard to 0.2). Similarly, mercurycontrol costs are greater on a per ton basis when we introduce the more stringent mercury standard inOption 1c. These findings are consistent with the basic economic principle of increasing marginal cost.Specifically, the cost of controlling increasing amounts of a pollutant will rise as additional reductionbecome more difficult technologically and economically.

Exhibit 5-3 also provides an estimate of cost per ton reduced across all HAPs in the rule. Asshown, costs for all facilities considered together range from about $5,000 (in Option 4) to about $8,500(in Option 2b). In general, the cost across all HAPs is lower for incinerators than for cement kilns and2

LWAKs.

EXHIBIT 5-3COST PER UNIT OF HAP REDUCED

ALL FACILITIES

432B2A1C1BOriginal FloorUnit

$54,583$52,708$51,884$54,029$59,922$61,072$40,581$/gD/F$8,821,427$9,212,333$7,952,142$7,602,764$7,394,788$3,618,479$3,711,855$/tonMercury

$297,325$299,420$304,959$281,773$325,032$319,931$337,445$/tonLVM$230,823$227,579$233,577$243,889$218,834$296,255$321,732$/tonSVM

$22,825$22,098$22,816$22,980$23,992$21,577$20,753$/tonChlorine$3,173$3,017$7,771$8,212$6,099$7,418$7,171$/tonHCl

NANANANANANANA$/tonTotal Chlorine$3,223$3,142$3,141$3,230$3,194$3,340$3,399$/tonParticulate$2,008$1,031$1,472$1,991$1,028$1,034$1,158$/tonCO

$38,447$795$8,583$38,883$812$779$557$/tonTHC

TOTAL ACROSS$5,220$6,914$8,576$5,389$7,743$6,309$5,564$/tonALL HAPs

CEMENT KILNS


$25,775$23,844$24,584$26,508$26,013$26,102$24,617$/gD/F$7,165,085$8,011,037$6,692,959$5,954,064$6,511,581$2,616,174$1,880,694$/tonMercury$1,959,332$1,955,586$1,984,893$2,005,034$2,207,708$2,522,016$2,549,407$/tonLVM

$350,494$299,265$337,140$383,234$330,856$473,196$543,140$/tonSVM$64,102$61,020$61,428$64,032$65,090$59,068$54,461$/tonChlorine$11,475$12,427$12,301$13,772$10,126$10,917$10,595$/tonHCl

NANANANANANANA$/tonTotal Chlorine$2,397$2,218$1,369$1,483$1,408$1,708$1,840$/tonParticulate$2,090NANA$2,094NANANA$/tonCO

$34,074$0$0$34,018$0$0$0$/tonTHC


LWAKs


NANANANANANANA$/gD/F$20,752,535$18,662,141$18,343,578$18,343,578$17,456,175$6,169,987$6,169,987$/tonMercury

$3,595,345$3,797,543$4,832,486$4,832,486$5,395,744$5,010,666$5,010,666$/tonLVM$1,809,622$3,840,516$2,244,050$2,244,050$2,086,975$3,686,954$3,686,954$/tonSVM


NANANANANANANA$/tonTotal Chlorine$21,119$48,344$22,630$22,630$16,069$19,780$19,780$/tonParticulate

NANANANANANANA$/tonCO$575,138NA$581,829$581,829NANANA$/tonTHC


Notes:Available data suggest that no reduction in D/F emissions is needed at LWAKs.

INCINERATORS


$389,399$388,174$369,170$373,877$454,014$467,505$234,442$/gD/F$9,952,900$9,977,540$8,901,085$9,042,768$7,869,247$4,627,959$5,095,076$/tonMercury

$187,915$189,570$189,717$163,923$195,894$173,717$190,726$/tonLVM$146,741$145,242$153,091$144,044$135,447$159,272$160,842$/tonSVM


NANANANANANANA$/tonTotal Chlorine$4,763$4,785$4,926$4,989$5,014$4,989$4,971$/tonParticulate$1,591$1,031$1,472$1,472$1,028$1,034$1,158$/tonCO

$158,705$103,349$190,327$193,073$105,570$101,314$40,675$/tonTHC


Note:Total annual compliance costs by HAP used to calculate the cost per unit of HAP reduced do not include CEM and permitting costs.1.

Estimated using Table 3 of Federal Register, Vol. 59, No. 181, September 20, 1994. Total tonnage3

reduced is 133,400 tons with total estimated annual costs of $445 million ($445 million/133,400 = $3,336).Inflated to 1994 dollars using GDP implicit price deflator.

Estimated using Table 9b of Federal Register, Vol. 60, No. 38, February 27, 1995. Total tonnage4

reduced is 74,080 tons with total estimated annual costs of $351 million (see p. 10667).


For context, it is useful to compare this cost-effectiveness measure with that estimated for otherair pollution regulations. Exhibit 5-4 presents information on the cost-effectiveness of other combustionregulations. As shown, the cost-effectiveness of the hazardous waste MACT standards are roughly on parwith the cost-effectiveness estimated for the 1994 Municipal Waste Combustor rule and the recentlyproposed rules for medical waste incinerators. This outcome is likely due to the fact that the MWC ruleincluded many of the same HAPs as the hazardous waste MACT rule (e.g., HCl, PM, metals, dioxin,mercury); however, the comparison is not direct (e.g., the MWC rule covers SO2), reducing somewhat thevalidity of the comparison. Note that the comparison is more complex when we consider individual facilitytypes. While incinerator cost-effectiveness compares favorably with other rules ($5,700 to $8,300 undervarious options), the cost-effectiveness for kilns, particularly LWAKs, significantly exceeds that of theother rules.

Exhibit 5-4

COMPARISON OF COST-EFFECTIVENESS

Rule ($ per ton reduced)

EstimatedCost-Effectiveness

MACT Standards for Hazardous WasteCombustors $5,000 - $8,500

MWC Proposed 1994 Subpart CbGuidelines (September, 1994) $3,9033

Medical Waste Incinerators ProposedRule (February, 1995) $4,7384

5.2 HEALTH BENEFITS

The health benefits sections below discuss the pathways and transport of dioxin and mercury thatresult in human exposure and the health impacts of dioxin and mercury.

U.S. Environmental Protection Agency, Estimating Exposure to Dioxin-Like Compounds, Volume I,5

June 1994.

U.S. Environmental Protection Agency, Health Assessment Document for6

2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds, Volume II , June 1994.

U.S. Environmental Protection Agency, Health Assessment Document for7

2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds, Volume III , August 1994.

Ibid.8

Ibid.9


5.2.1 Exposure to Dioxin Emissions and Health Impacts

Polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans, hereafter referred tocollectively as dioxins, are ubiquitous in the environment. The more highly chlorinated dioxins, which areextremely stable under environmental conditions, persist in the environment for decades and are foundparticularly in soils, sediments, and foods. It has been hypothesized that the primary mechanism by whichdioxins enter the terrestrial food chain is through atmospheric deposition. Dioxins may be emitted directly5

to the atmosphere by a variety of anthropogenic sources or indirectly through volatilization or particleresuspension from reservoir sources such as soils, sediments, and vegetation.

The most well known incident of environmental contamination with dioxins occurred in Seveso,Italy in an industrial accident. Symptoms of acute exposures such as chloracne occurred immediatelyfollowing the incident. Since then, significant increases in certain types of cancers have also beenobserved. After evaluating a variety of carcinogenicity studies in human populations and laboratory6

animals, EPA has concluded that 2,3,7,8-tetrachlorodibenzo-p-dioxin and related compounds are probablehuman carcinogens. EPA estimates that a dose of 0.01 picograms on a toxicity equivalent (TEQ) basis7

per kilogram body weight per day is associated with a plausible upper bound lifetime excess cancer riskof one in one million (1 x 10-6). Toxicity equivalence is based on the premise that a series of common8

biological steps are necessary for most if not all of the observed effects, including cancer, from exposuresto 2,3,7,8 chlorine-substituted dibenzo-p-dioxin and dibenzofuran compounds in vertebrates, includinghumans. Given the levels of background TEQ exposures discussed below, as many as 600 cancer casesmay be attributable to dioxin exposures each year in the United States.

EPA has also concluded that there is adequate evidence from both human populations andlaboratory animals, as well as other experimental data, to support the inference that humans are likely torespond with a broad spectrum of non-cancer effects from exposure to dioxins if exposures are highenough. Although it is not possible given existing information to state exactly how or at what levelsexposed humans will respond, the margin of exposure between background TEQ levels and levels whereeffects are detectable in humans is considerably smaller than previously thought.9

U.S. Environmental Protection Agency, Estimating Exposure to Dioxin-Like Compounds, Volume II ,10

June 1994.

Ibid.11

U.S. Environmental Protection Agency, National Study of Chemical Residues in Fish, Office of12

Science and Technology, September 1992.

U.S. Environmental Protection Agency, Estimating Exposure to Dioxin-Like Compounds, Volume II ,13

June 1994.

Other major dioxin emitters include medical waste incinerators and municipal waste incinerators.14

Yearly dioxin emission estimates for these sources are 5.1 kg TEQ and 3.0 kg TEQ, respectively.


Dioxins are commonly found in food produced for human consumption. Consumption of dioxincontaminated food is considered the primary route of exposure in the general population. EPA evaluateddata collected in four U.S. studies, three of which included analyses of all 2,3,7,8 chlorine-substitutedcongeners of dibenzo-p-dioxin and dibenzofuran. EPA's evaluation concluded that "background" levelsin beef, milk, pork, chicken, and eggs are approximately 0.5, 0.07, 0.3, 0.2, and 0.1 parts per trillion freshweight, respectively, on a toxicity equivalent (TEQ) basis. EPA then used these background levels,10

together with information on food consumption, to estimate dietary intake in the general population. Thatestimate is 120 picograms TEQ per day.11

EPA has also collected data on dioxins in fish taken from 388 locations nationwide and found thatat 89 percent of the locations, fish contained detectable levels of at least two of the dioxin and furancompounds for which analyses were conducted. (Of the 2,3,7,8 chlorine-substituted congeners, only12

octachlorodibenzo-p-dioxin and octachlorodibenzofuran were not analyzed.) Seven of the compounds,including 2,3,7,8-TCDD, were detected at over half the locations. Detection limits were generally at orbelow 1 part per trillion on a toxicity equivalent basis. The median (50th percentile) concentration in fishon a toxicity equivalent basis (TEQ) was 3 parts per trillion (ppt) while the 90th percentile wasapproximately 30 ppt TEQ. Five percent of the sites exceeded 50 ppt TEQ. At most sites, both acomposite sample of bottom feeders and a composite sample of game fish were collected. At sitesconsidered representative of background levels, the median concentration was 0.5 ppt TEQ.

EPA has estimated that hazardous waste incinerators and hazardous waste burning cement andlightweight aggregate kilns currently emit 0.08, 0.86, and less than 0.01 kg TEQ of dioxins per year,respectively, or a total of 0.94 kg TEQ per year. Excluding non-hazardous waste burning cement kilns, anemission rate of approximately 9 kg TEQ per year is estimated for all other U.S. sources. Therefore,13

hazardous waste burning sources represent about 9 percent of total anthropogenic emissions of dioxins inthe U.S. 14

There is information to suggest, however, that dioxin emissions are higher than have beenestimated. Public comments on EPA's dioxin reassessment have identified a number of possible additionalsources of dioxins, including decomposition of materials containing chlorophenols (i.e. wood treated withPCP), metals processing industries, diesel fuel and unleaded gasoline, PCB manufacturing, andre-entrainment of reservoir sources. Reservoir sources may be a significant source of vapor phase dioxins.On the other hand, emissions from at least one of the sources, medical waste incinerators, is probably


significantly overestimated. Supporting the view that dioxin emissions may be higher than previouslyestimated are indications that deposition may be considerably greater than can be accounted for bypresently identified emissions.

The impact of emissions on exposure and risk depends on the relative geographic locations of theemission sources and receptors which contribute to exposure and risk, primarily farm animals. This appliesto both near field dispersion and long-range transport and it affects exposure and risk both in determiningwhether the trajectory of an air parcel impacts receptors of concern and in determining the chemical fateof the emissions. The fate of dioxins depends on degradation processes that can occur in the atmosphere.These processes can increase or decrease the toxicity of the original emissions through dechlorination. Thisprocess can have different effects on different emission sources, depending on the congener distributions,residence time in the atmosphere, and climatic conditions.

Considering all these factors, it is apparent that hazardous waste burning sources contributesignificantly to the overall loading of dioxins to the environment, although the relative magnitude of thecontribution remains to be determined. Similarly, it may be inferred that hazardous waste burning sourcescontribute significantly to dioxin levels in foods used for human consumption and, to an extent as yetunknown, the estimated 600 cancer cases attributable to dioxin exposures annually.

EPA estimates that dioxin emissions from hazardous waste burning sources will be reduced toapproximately 0.03 kg TEQ per year under the MACT standards. These reductions would result in adecrease of between 8 and 9 percent in total estimated anthropogenic U.S. emissions. EPA expects thatreductions in dioxin emissions from hazardous waste burning sources, in conjunction with reductions inemissions from other dioxin-emitting sources, will help reduce dioxin levels over time in foods used forhuman consumption and, therefore, reduce the likelihood of adverse health effects, including cancer,occurring in the general population.

5.2.2 Exposure to Mercury Emissions and Health Impacts

Mercury has long been a concern in both occupational and environmental settings. The mostbioavailable form of mercury and, therefore, the form most likely to have an adverse effect, is methylmercury. Human exposures to methyl mercury occur primarily from ingestion of fish. As a result ofmercury contamination, there are currently fish consumption bans or advisories in effect for at least onewaterbody in over two thirds of the States.

Nationally, about 60 percent of all fish consumption bans and advisories are due to mercury. Inseveral States the mercury advisories are statewide, with the most widespread concerns being in thenorthern Great Lakes states and Florida. The bans and advisories vary from state to state with respect tothe levels of concern, the recommended limits on consumption, and other factors. Therefore, it is difficultto develop a national estimate of potential risk based on this information. Nevertheless, these bans andadvisories provide one indication of the extent and severity of mercury contamination.

Even low levels of mercury in surface waters can lead to high levels of mercury in fish. EPA hasestimated that bioaccumulation factors, which represent the ratio of the total mercury concentration in fishtissue to the total concentration in filtered water, range from 5000 to 10,000,000 depending on the speciesof fish, the age of the fish, and the waterbody the fish inhabit.

U.S. Environmental Protection Agency, Mercury Study Report to Congress Volume VI:15

Characterization of Human Health and Wildlife Risks from Anthropogenic Mercury Emissions in the UnitedStates, December 1994.

U.S. Environmental Protection Agency, National Study of Chemical Residues in Fish,16

Office of Science and Technology, September 1992.

U.S. Environmental Protection Agency, Mercury Study Report to Congress Volume III: An17

Assessment of Exposure from Anthropogenic Mercury Emissions in the United States , December 1994.


The most well known example of mercury poisoning from ingestion of fish occurred in the vicinityof Minamata Bay, Japan. Severe neurological effects resembling cerebral palsey occurred in the offspringof exposed pregnant women. EPA has estimated what it considers a safe level of exposure to methylmercury. This level, referred to as the reference dose, is 1E-4 mg/kg-day. The reference dose is based onan evaluation of 81 maternal-infant pairs exposed to methyl mercury in an incident in Iraq in which methylmercury treated seed grain was diverted for use in making bread. The reference dose, which is based ondevelopmental effects in children, is also considered to be protective with respect to neurological effectsin adults (e.g., paresthesia). An uncertainty analysis of the Iraqi data concluded that the reference dose fallsbelow the 5th percentile of the threshold for developmental neurologic abnormalities in human infants and,therefore, represents a lower bound estimate of the threshold. A major premise of the analysis is that the15

81 pregnant Iraqi women are representative of the most susceptible subgroup in the general population(e.g., in the United States). Other sources of uncertainty are the duration of the maternal exposure(approximately three months), latency in the appearance of effects (from as little as a month to as long asa year), possible misclassification of maternal exposures, differences in the vehicle of exposure (i.e., grainvs. fish), and the selection of neurologic or behavioral endpoints.

EPA collected data on chemical residues in fish taken from 388 locations nationwide and foundthat at 92 percent of the locations, fish contained detectable levels of mercury. (Detection limits varied16

between 0.001 and 0.05 parts per million.) The median (50th percentile) mercury concentration in fish was0.2 ppm while the 90th percentile was 0.6 ppm. Two percent of the sites exceeded 1 ppm. At most sites,both a composite sample of bottom feeders and a composite sample of game fish were collected. Thehighest concentration, 1.8 ppm, was measured at a remote site considered to represent backgroundconditions.

Similar results have been obtained in other studies, strongly suggesting that long-range atmospherictransport and deposition of anthropogenic emissions is occurring. This conclusion is further supported bya modeling analysis which found that there is no region in the continental U.S. where deposition of mercuryis not occurring. Therefore, it is likely that mercury emissions contribute to both regional and global17

deposition, as well as deposition locally, and subsequent impacts on surface waters.

An indication of the significance of mercury contamination in fish is illustrated by combining dataon the levels of mercury in fish with data on fish consumption and comparing it to the reference dose formethyl mercury. For example, a fish consumption rate of 140 g/day (a 90th percentile rate associated withrecreational fishing) in conjunction with a mercury concentration of 0.6 µg/g (a 90th percentileconcentration) translates into an average daily dose of 1E-3 mg/kg-day, or 10 times the reference dose.Using the same fish concentration with a mean fish consumption rate for recreational anglers of 30 g/daygives a dose that is three times the reference dose. At the median fish concentration of 0.2 µg/g and a fish

U.S. Environmental Protection Agency, Mercury Study Report to Congress Volume II: Inventory of18

Anthropogenic Mercury Emissions in the United States, December 1994.

Other major mercury emitters include medical waste incinerators, municipal waste incinerators, utility19

boilers, and commercial/industrial boilers. Mercury emission estimates for these sources are 58.8, 57.7,47.8, and 26.2 Mg/year, respectively.


consumption rate of 30 g/day, the dose is nearly 90 percent of the reference dose. These results indicatethat for persons who eat significant amounts of freshwater fish, exposures to mercury are significant whencompared with EPA's estimate of the threshold at which effects may occur in susceptible individuals.However, it must be recognized that EPA's threshold estimate represents a lower bound; the true thresholdmay be higher than EPA's estimate.

EPA has estimated that hazardous waste incinerators and hazardous waste burning cement andlightweight aggregate kilns currently emit 4.2, 5.6, and 0.3 Mg of mercury per year, respectively, or a totalof 10.1 Mg per year. An emission rate of 230 Mg per year has been estimated for all other U.S. sources.18

Therefore, hazardous waste burning sources represent about 4 percent of total anthropogenic emissions ofmercury in the U.S. From these estimates it is apparent that hazardous waste burning sources contribute19

significantly to the overall loading of mercury to the environment and, by inference, to mercury levels infish.

EPA estimates that mercury emissions from hazardous waste burning sources will be reduced to2.8 Mg per year at the floor level and to 1.9 Mg per year at the proposed above the floor standard. Thesereductions would result in reductions of total anthropogenic U.S. emissions of approximately 3 percent.EPA expects that reductions in mercury emissions from hazardous waste burning sources, in conjunctionwith reductions in emissions from other mercury-emitting sources, will help reduce mercury levels in fishover time and, therefore, reduce the likelihood of adverse health effects occurring in fish-consumingpopulations.

5.3 ECOLOGICAL RISKS: METHODOLOGY AND RESULTS

5.3.1 Methodology for Assessing Ecological Risks

Emissions from waste burning may also affect terrestrial and aquatic ecosystems in the areasaround combustion facilities. For example, deposition of air pollutants may affect the food web of thesurrounding ecosystem in subtle ways that undermine the viability of the ecosystem. Ecological riskassessments should be especially attentive to persistent pollutants that bioaccumulate in plants and animals(e.g., metals) given that these are the pollutants that are not destroyed in the combustion process.

As an initial screen for assessing risk to aquatic ecosystems, EPA compared the concentration ofpollutants in surface water to EPA-approved aquatic toxicity criteria. Phone surveys were used to select

A more detailed explanation of the ecological risk method can be found in Risk Assessment Support20

to the Development of Technical Standards for Emissions from Combustion Units Burning Hazardou sWastes: Background Information Document, draft, prepared for Industrial Economics, Incorporated,prepared by Research Triangle Institute, August, 1995.


waterbodies within 20 kilometers of the facility. The change in pollutant concentrations in the water bodieswas calculated as a function of deposition rates and the volumetric flow rates or depth of the water body.20

5.3.2 Ecological Risk Results

For the watersheds examined at sample facilities, dioxin exhibits the potential for causingexceedences of ecological risk criteria in surface water. As shown in Exhibit 5-5, central tendency (50thpercentile) dioxin emissions from cement kilns may cause exceedences in the most affected watersheds.High end emissions of dioxins (90th percentile) from both incinerators and cement kilns are estimated tocause exceedences of the criterion.

The analysis of ecological risk suggests that water quality criteria may be exceeded in the mostvulnerable watersheds around waste combustion facilities, particularly cement kilns. This generally occurs,however, only under high end emission assumptions. The conservative nature of the analysis is alsoreflected in the fact that the criteria exceeded are based on exposures to species higher in the food chainthat feed exclusively on fish from the contaminated water bodies, further reducing the likelihood ofecological effects.

5.4 OTHER SOCIO-ECONOMIC BENEFITS

In addition to the health and ecological benefits described above, the proposed rule for hazardouswaste combustion facilities may result in a variety of other economic benefits to society. Most importantly,aesthetic and health risk disamenities may be reflected in a decrease in property values around combustionfacilities. The discussion below summarizes EPA's estimates of the potential magnitude of these propertyvalue effects. In addition, we qualitatively discuss several other benefit categories that may be influencedby the MACT standards, including materials damage and soiling of buildings, aesthetic nuisances (e.g.,noise, odor), and recreational and commercial fishing opportunities.


Exhibit 5-5

SUMMARY OF BASELINE ECOLOGICAL RISK RESULTS

Facility Type

Ratio of Total Water Concentration toWater Quality Criteria for Dioxin

CentralTendency High End

Incinerators 0.008 - 0.4 0.2 - 13

Cement Kilns 0.1 - 9.0 1.0 - 97

LWAKs 0.003 - 0.2 0.003 - 0.15

Note: Dioxin ratios are based on total water concentrations.

5.4.1 Property Value Benefits

EPA performed a screening analysis to assess the potential magnitude of property value effectscaused by the presence of hazardous waste combustors. The underlying hypothesis is that residentialproperty values will reflect not only standard attributes (e.g., house size, city services), but also thepresence of environmental disamenities. As discussed below, we performed a literature review to evaluateprevious studies of the link between property values and environmental problems such as landfills andincinerators. Using results from the most relevant of these studies, we develop a benefits transfer analysisthat employs assumptions that provide conservative, low-end estimates of property value benefitsassociated with both the closure of commercial incinerators as well as the reduction in emissions from othercombustors.

5.4.1.1 Background and Summary of Existing Studies

Researchers have applied several techniques to estimate the impact of a disamenity, such as a wastesite, on property values; these include real estate appraiser questionnaires, tax assessor surveys, and hedonicprice models. Hedonic price models generally produce the most reliable estimates because they rely onempirical data and employ relatively rigorous modeling approaches. A well-designed hedonic modelisolates the effect of the environmental disamenity on housing prices by controlling for other attributes ofthe house and for neighborhood characteristics. The available hedonic models are quite diverse in theirspecification, and rely on a variety of estimation techniques and data sources. Thus, model results mustbe interpreted with care.

A literature search was conducted to assess the range of property value impact estimates and toexamine some of the differences among estimates. The ideal study would be one that analyzes propertyvalue impacts around a hazardous waste combustor. Such a study would also examine the effect of reducedemissions on property values. While no such ideal study exists, we did identify ten relatively recent studiesthat used hedonic modeling concepts to estimate property value impacts of environmental disamenities.


A list of these ten studies and a summary of key results is found in Exhibit 5-6. None of the availablestudies considered hazardous waste combustors. One study examined property value effects fromincinerator operations; however, this facility burned only municipal wastes. The remaining studies can begrouped into three categories on the basis of site analyzed -- municipal landfills, hazardous waste landfills,and hazardous waste sites (including Superfund sites).

While these studies all used hedonic modeling concepts to measure property value impacts, thestudies differed in their objectives. Some sought to estimate property value impacts as a function ofdistance from the disamenity. The objective of other studies was to determine sale price effects as afunction of the number of disamenities (e.g., the number of hazardous waste sites within a given locality).Since the objectives of the studies were often different, care must be taken in comparing studies.Additionally, since the ideal hazardous waste combustor study was not found, the results in the studiessummarized below cannot be directly applied to the issue of hazardouswaste combustion. The BenefitEstimates section explains how previous study results were used to develop property value benefit estimatesfor hazardous waste combustors.

The key element drawn from the studies is the estimated marginal impact that distance from thesite has on property values. This marginal impact will be referred to as the premium, in dollars per mileper household. For example, assume that the distance premium per household is $5,000 per mile. Thepremium indicates that a home's value will increase by $5,000 for each mile further from the site. Thesections below discuss the literature review findings with respect to this distance premium.

Incinerator Study

The recent study by Kiel and McClain (1995) analyzes the impact of a municipal waste incineratoron housing prices. Using data covering the time before the incinerator was even proposed through the timeof ongoing incinerator operations, they show that price impacts are not constant over time. Price impactsappear strongest during the time when the incinerator first began operating. The premium during this timewas $8,100 per mile per household. This contrasts with a premium of $2,283 during the constructionperiod, and a premium of $6,607 during ongoing operations, seven years after the incinerator beganoperating. The price response to site proposal and publicity (the rumor stage) was weak. Housing pricesincrease with distance from the site up to approximately 3.5 miles.


Exhibit 5-6SUMMARY OF HEDONIC PROPERTY VALUE STUDIES

Study Years of Data Location Type of Premium DistanceFacility ($ per mile per household) Range

(House Sales Price)

Smolen (1992) 1986 through Toledo, Ohio Hazardous Envirosafe Landfill: mid 1990 Waste $9301 - $14,205 0-2.6 miles

Landfills $7416 - $22,793 2.6-5.75 miles

Riga Landfill:$644 - $4640 2.6-5.75 miles

Kiel and McClain 1974-1992 North Incinerator Construction: Effect(1995) Andover, $2283/mile becomes

Mass. Online: negligible at $8100/mile 3.5 miles.Ongoing Operations: $6670/mile

Kohlhase (1991) 1976, Houston, Hazardous $3260 per mile 0 - 6.2 miles 1980, Texas Waste Sites (1985)1985 Insig. for 1976 and 1980

Nelson, 1979-1989 Ramsey, Landfill $4896 per mile 0 - 2 milesGenereux, MinnesotaGenereux (1992)

Mendelsohn, 1969-1988 New Bedford, PCB Pollution $9000 for zone 1 Within 2Hellerstein, Mass. in the Harbor (Inner Harbor) miles.Huguenin,Unsworth, Brazee $7000 for zone 2(1992) (Outer Harbor)

McClelland, 1983-1985 Los Angeles Hazardous House Price reduced by No distanceSchulze, Hurd Waste Landfill $2084 per 10% increase in range. (1990) the proportion of high risk

neighborhood respondents

Smith and 1984 Suburban Hazardous Consumer surplus estimate of No distanceDesvousges Boston Waste Landfill $330 to $495 annually. range.(1986)

Bleich, Findlay, 1978-1988 Los Angeles Landfill No significant property value Within onePhillips (1991) effects found. mile of the

landfill.

Ketkar (1992) 1980 New Jersey Hazardous $1300-$2000 increase in the PremiumWaste Sites median property value for applies to

one fewer HW site in the homes in amunicipality single

municipality.

Michaels and 1977-1981 Suburban Hazardous Full Sample: $115 Homes locatedSmith (1990) Boston Waste Sites Premier Market: 461 within 13

Above Average: 58 miles of theAverage Market: 48 nearest

(annualized at 10%) waste site.hazardous

Note that the distance premiums reported have not been adjusted to 1994 dollars. In most cases the21

articles do not specify the dollar year for the figures reported.

These results were annualized assuming a 10 percent discount rate. Total present value premiums22

would be on the order of $4,610, $580, and $480 for the premium, above average, and average markets,respectively.

Response rate of 45 percent.23

The total present value premium per mile would be roughly $3,300 to $4,950 per mile per household,24

assuming a 10 percent discount rate.


Hazardous Waste Sites

Four of the studies assessed the impact of hazardous waste sites on housing prices. Some of thesites considered are on the National Priorities List. Ketkar (1992) estimated a $1,300 to $2,000 increasein the median property value for one fewer hazardous waste site in the municipality. Kohlhase (1991)21

estimated a premium of $2,360 per mile per household for homes located within 6.2 miles of the site. Thispremium was significant only after the site was placed on the NPL. The primary finding was that housingprices were not affected by the presence of the hazardous waste site but rather by the EPA announcementregarding which sites would be placed on the NPL. The study by Michaels and Smith (1990) concludedthat different neighborhood submarkets (premier, above average, or average) result in significantlydifferent distance premiums. The premier market commanded an annualized premium of $461 per mileper household, the above average market had a premium of $58 per mile, and the average submarket hadan associated premium of $48 per mile. 22

The fourth study analyzed the price effects of a PCB-contaminated harbor (Mendelsohn, et. al.,1992). The harbor study analyzed repeat sales data to determine the property value impacts of PCBpollution in the New Bedford harbor sediments. Their results indicate that housing prices decreased by 8percent (median of $9,000) for homes that are closest to the inner harbor and a decrease of 7 percent($7,000) for homes located near the outer harbor.

Hazardous Waste Landfills

All three studies that looked at hazardous waste landfills concluded that these sites adversely affectprices of nearby homes. The premium estimated by Smolen (1992) was $12,061 for homes within the 2.6mile range, and $12,106 for homes within 2.6 and 5.75 miles of the site. McClelland et. al (1990) assessedthe decrease in home prices as a function of neighborhood risk perception. Neighborhood risk wasdetermined through a survey mailed to 1912 addresses. The proportion of responses that fell into the high23

risk group was calculated for each neighborhood. The neighborhood risk proportion was calculated bothbefore and after the landfill closure, and the appropriate risk value was attached to the sales price dependingupon the sale date. They found an estimated decrease of $2,084 in home sales price per 10 percent increasein the proportion of neighborhood respondents in the high risk group. The third study by Smith andDesvousges (1986) estimated the premium through survey questions, developed using the conceptualframework of a hedonic property value model. A demand for distance model was subsequently estimatedusing the survey results; consumer surplus was estimated at between $330 and $495 per year per mile froma disposal site.24

Other EPA studies have employed benefits transfer to estimate property value benefits. For example,25

see Benefit and Net Benefit Analysis of Alternative National Ambient Air Quality Standards For ParticulateMatter, prepared for Office of Air Quality Planning and Standards, March 1983, Volume III.


Municipal Landfills

Two studies evaluated the impact on property values from municipal landfills. One studyconcluded that a "well-designed and well-managed" landfill appears to have no statistically significantimpact on surrounding property values. (Bleich, et. al. 1991) The other study concluded that a municipallandfill in Minnesota did, in fact, cause decreases in property values to a distance of about two miles. Theestimated property value premium was $4,896 per mile per household, over a range of two miles from thesite. (Nelson et. al. 1992)

5.4.1.2 Benefits Transfer Analysis: Methodology and Obstacles

Property value impact estimates are often desired by EPA and other regulatory agencies, but timeand financial constraints often prohibit performing original research. To develop an estimate of propertyvalue effects near hazardous waste combustors, a benefit transfer methodology is employed. Benefit25

transfer analysis involves applying results from existing studies to the situation of interest. The accuracyof estimates from this method depends upon the similarity of situations considered. In this section, theapplicability of results from the studies summarized above to waste burning is discussed. Specifically,estimates of property value impacts resulting from reduced emissions at facilities that burn hazardouswastes and cessation of burning hazardous wastes at various facilities are desired.

Several issues and problems must be addressed in developing a benefit transfer analysis to estimatethe effect of changes in waste burning behavior. First, most studies estimate property value impactsresulting from a total land-use disamenity (e.g., the presence of a waste site). In contrast, the proposed ruleproduces reduced emissions or cessation of waste burning at large, multi-purpose industrial facilities. Itis difficult to determine the extent to which individual factors of a facility are contributing to propertydevaluation. The mere existence of industrial activity may be a disamenity. Smoke and other emissionsare aesthetically displeasing and may produce unpleasant odors. Similarly, increased truck traffic and theassociated noise and air pollution also are factors which contribute to depressed property values. Thehazardous waste combustion MACT standards will generally only influence one of the factors at thefacility, i.e. emission rates. Also, some units will discontinue burning hazardous wastes. But even at thesefacilities, the primary industrial activities (e.g., cement manufacturing) will remain in operation. Thus,property value effects may be limited relative to studies that measure the influence of an entire facility(e.g., landfill).

Second, while results from air pollution property value studies could be used to determine theproperty value benefit of reduced emissions at combustion facilities, existing air pollution studies are notadequate for the following reasons:

!! The pollutants analyzed in existing studies are not the same as the ten hazardousair pollutants regulated under the proposed MACT standards.

Kerry Smith and Ju-Chin Huang, "Can Markets Value Air Quality? A Meta-Analysis of Hedonic26

Property Value Models," Journal of Political Economy, Volume 103 (1) 209-227, 1995.

Ibid, 389.27

The study reported premiums for five different phases of operation: the pre-rumor stage (the period28

before the incinerator was ever proposed), the rumor stage (the period when the incinerator was proposed,but before any construction), the construction period, the online stage (a period of three years, from thetime the incinerator began operations), and the ongoing operation stage (four to seven years after theincinerator began operations). The premium estimated for the ongoing operation stage, $6,670, was usedbecause most of the combustion facilities have been operating for at least four years.

Cement kilns, LWAKs, and facilities with on-site incinerators will continue their primary operations29

even if the facility stops burning hazardous wastes. Commercial incinerators are the only units expectedto cease operations altogether when they stop burning hazardous wastes.


!! Property value effects due to air pollution have been found to depend critically onthe market conditions of a given location.26

!! These studies analyze old data (1960-1978). A comprehensive review of hedonicmodels and air pollution by Smith and Huang reports that "the most current data27

set used in a published hedonic model with air pollution evaluated for a U.S. citywas 1980."

Finally, premiums may vary between combustion facilities, depending upon facility type,geographic location, baseline emissions, and other location-specific market conditions. For example,baseline emissions data for hazardous air pollutants from combustion suggest that emissions percombustion unit vary widely, and that there could be different HAPs of concern in different combustionsectors.

5.4.1.3 Benefit Transfer Estimates

The issues described above demonstrate the complexities associated with direct extrapolation ofresults from existing studies. Benefits transfer analysis, however, is possible as long as key parameters arecarefully addressed. Specifically, two key questions must be answered before a benefit estimate can bederived: (1) Which premium should be applied to homes? and (2) What distance range should be used?A reasonable premium and distance range can be selected from a study which analyzes a site withcharacteristics similar to combustion facilities. The most comparable site is the municipal wasteincinerator analyzed by Kiel and McClain (1995). The per-household premium averages $6,670 per mileand effects are significant out to a distance of 3.5 miles from the site. Using this premium and distance28

range could result in understated benefits if the hazardous waste combustion facilities affected by thisproposed rule are viewed as more of a threat than the municipal waste incinerator studied by Kiel andMcClain. On the other hand, benefits would be overstated if weaker impacts are likely for facilities thatreduce emissions, relative to facilities that completely cease operations. The approach used to evaluate29

the range of potential property value impacts of the proposed MACT standards is described below.

These same counties were used in the health risk assessment for this rule.30

The Kiel and McClain premium was adjusted by multiplying it by the ratio of the median house price31

in the county in which the sample facility is located to the median house value in Kiel and McClain.

Population density at a more refined scale would improve the analysis because it would more32

accurately portray the distribution of homes around a site. EPA has performed a GIS analysis that providesmore precise population and housing density information. While time was not available to incorporatethese data, they will be used in future refinement of this analysis.


The basis for this assessment is an analysis of property value impacts at a sample of ten combustionfacilities. The premium from Kiel and McClain was adjusted for each of the ten counties in which these30

facilities are located to reflect differences in median house values. These adjusted premiums are shown31

in Exhibit 5-7; all are lower than the premium estimated by Kiel and McClain because home prices in theneighborhood analyzed by Kiel and McClain (North Andover, Massachusetts) are relatively high. Housingdensities for the ten counties were calculated at the county level by dividing total occupied households percounty by the county land area found in 1990 Census data. The estimates in EPA's analysis assume an evendistribution of homes within each county.32

The key issue in determining the effects of the proposed rule on property values is how to valuethe impacts of reductions in concentrations of air pollutants around facilities that continue to burn waste.Past studies such as Kiel and McClain have looked at how the presence of a facility affects property values,not at the impact of changes in emissions levels. In reality, the new MACT standards probably do not fullyeliminate all property value impacts resulting from location of homes near hazardous waste combustionfacilities. In this case, we need to develop a benefits transfer approach that considers emissions reductions.One possible approach for getting a very rough handle on the impact of pollutant concentration reductionsis to view increases in distance from a combustion facility as a proxy for reductions in concentration. Forexample, after emissions are reduced, a home located one mile from a combustion facility may be exposedto pollutant concentrations similar to pre-rule concentrations at homes two miles from the facility. In thiscase, the benefit of the rule is an increase in the value of the home by the difference in housing pricesbetween one and two miles (i.e., the increase in value for that home is equal to the per mile premium).

The distance-proxy estimation approach requires three key input parameters: (1) the per-milepremium, (2) the boundary distance within which the premium applies, and (3) the mileage used to proxyemissions reduction (mileage proxy). The boundary distance and the mileage proxy have the greatestimpact on the benefit estimates because benefits increase geometrically with distance; the impact ofchanges over the range of premiums suggested by other studies is much more limited.

Unfortunately, because no studies have ever looked at distance as a proxy for concentrationreductions at hazardous waste facilities, all three of these input values are highly uncertain. As a result,EPA has chosen to present an analysis that shows how property value impacts would change as the keyparameters are varied over their expected ranges. For this analysis, benefit estimates were calculated undertwo boundary distance assumptions, one mile and 3.5 miles. For each of these boundary assumptions,property value impacts are graphed as a function of the effective increase in each home's distance from thesite that results from the reduction in concentration. To simplify the presentation, the premium is notvaried since its effect is much more limited. The results are shown in Exhibits 5-8 and 5-9.

EXHIBIT 5-7ADJUSTED DISTANCE PREMIUM PER HOUSEHOLD

AdjustedMedian PremiumHouse StateCounty

Value

$1,421$51,600IndianaPutnam$787$28,600KansasNeosho

$1,145$41,600MichiganAlpena$2,902$105,400PennsylvaniaNorthampton$3,054$110,900New YorkAlbany$1,390$50,500South CarolinaOrangeburg$6,041$219,400CaliforniaContra Costa$1,905$69,200LouisianaEast Baton Rouge$2,302$83,600MinnesotaRamsey$1,748$63,500TexasHarris

Property value benefit estimates reflect a variety of factors, including reductions in aesthetic33

nuisances and reductions in health risk to surrounding areas. Because of the latter, these benefits cannotsimply be added to the monetized benefits of cancer risk reduction; the degree of overlap may besignificant.


A specific example may make the implementation of the approach clearer. Consider a house thatis located one mile from a combustion facility. If the change in pollutant concentration resulting from theMACT standards is assumed to be equivalent to moving the house 1.5 miles further from the source, thenthe change in value is calculated as follows. First, the reduction in the value of a home located one milefrom the facility is calculated. This is equal to the premium ($6,670) times the number of miles the houseis from the boundary at which there is no longer any property value effect. In the case of the boundarybeing set at 3.5 miles, the decrease in property value is 2.5 times the premium ($16,675). Next, weestimate the property value reduction for a home that is 1.5 miles further from the site, in this case a houselocated at 2.5 miles. The reduction in value for this house is 1.0 times the premium ($6,670), since thehouse is located one mile from the point at which there is no longer any property value impact. Thedifference between $16,675 and $6,670 represents the increase in value due to the reduction in emissionsin our example. This calculation is performed for all houses around each of the ten combustion facilitiesin the sample. The values for each of the ten facilities are then averaged. The per facility average is thenmultiplied by the nationwide population of 222 combustion facilities to determine total national impacts.For each of the two assumed boundary distances (1.0 and 3.5 miles), impacts are calculated for a range ofconcentration distance proxies.

The analysis suggests that the range of potential property value impacts is quite wide and issensitive to both the boundary distance and the assumed mileage proxy for concentration reduction. If oneassumes that only homes within one mile of the combustion facility experience an increase in value, totalannual property value benefits range from $12 to $20 million as the distance proxy increases from 0.25 toone mile. Alternatively, assuming reductions in concentration affect all homes within 3.5 miles of thefacility, total annual property value impacts increase to between $200 and $840 million, depending on themileage proxy. The $840 million value is an effective upper bound in that it assumes all homes are33

moved outside the 3.5 mile range, at which point all the devaluation due to the combustion facilitydisappears. Overall, the analysis makes clear that increasing the boundary by 2.5 miles changes benefitestimates by more than an order of magnitude.

5.4.1.4 Conclusions

Review of available studies clearly indicates the potential for property value effects aroundhazardous waste combustion facilities. The types of sites and impacts analyzed in existing studies,however, are not directly analogous to the changes expected under the new MACT standards. In particular,the effect of the MACT standards differs from the effects reflected in available studies because: (1)emissions would be reduced rather than having a disamenity (e.g., landfill) fully removed; and (2) kilns andfacilities with on-site incinerators have other disamenities entirely separate from waste burning.

0

5

10

15

20 To

tal I

mpa

cts

(mill

ions

of d

olla

rs)

0 0.5 1 1.5 2 2.5 3 3.5 Mileage Proxy

(1 mile boundary)

Exhibit 5-8Total Annual Property Value Impacts

Note: No benefits after 1 mile.

0

200

400

600

800

1000

Tota

l Im

pact

s (m

illio

ns o

f dol

lars

)

0 0.5 1 1.5 2 2.5 3 3.5 Mileage Proxy

(3.5 mile boundary)

Exhibit 5-9Total Annual Property Value Impacts

In particular, a distance proxy more directly linked to the degree of emissions reduction provided by34

the MACT standard could be developed using air modeling information. At present, we are unable togauge the impact this additional analysis would have on the benefits estimates presented here. Furtheranalysis is also needed to determine the appropriate boundary within which property values are responsiveto changes in concentration.

Benefits Analysis of Alternative Secondary National Ambient Air Quality Standards for Sulfur Dioxide35

and Total Suspended Particulates, Mathtech, Inc., prepared for U.S. EPA, 1982.


In response to these complexities, EPA conducted an evaluation of potential benefits that usesdistance from a combustion facility as a proxy for emissions reduction. The analysis suggests that, ifdistance is an effective proxy for changes in concentration and if effects occur out to 3.5 miles from thesite, then only a very small increase in the effective distance of homes from the site yields property valuebenefits that exceed the costs of the rule. For example, assuming each house has concentrations ofpollutants reduced by an amount that makes it effectively 0.25 miles further from the site than its truedistance in the pre-regulatory baseline produces total annual benefits of nearly $200 million.

Unfortunately, to know whether these benefits really occur requires more knowledge about howhousing prices would respond to changes in air pollutant concentrations. While the use of distance as aproxy for emissions reduction has a certain commonsense appeal and provides some context for evaluatingwhether the new MACT standards could have significant impacts on property values, it does not answerthe key question -- whether property values are sensitive to changes in air pollution concentrations orwhether they are simply a function of the presence or absence of the facility itself. While the analysisdescribed above could be refined in a variety of ways, until this central question is answered, EPA is onlyable to conclude that property value impacts have the potential to be very significant. 34

5.4.2 Other Benefit Categories

The MACT standards may also provide a number of other types of economic benefits. Below, wediscuss potential effects on soiling and materials damages, aesthetic disamenities (e.g., noise, odor), andcommercial and recreational fishing opportunities. Note that these benefits may be partially reflected inthe property value effects reviewed above, and therefore cannot simply be added to property value benefits.

5.4.2.1 Soiling and Materials Damage

Hazardous waste combustion facilities release pollutants that have the potential to cause soiling andmaterials damages. Reductions in these emissions may result in decreases in society's expenditures oncleaning or repair of these damages. The most important combustor pollutant that might cause these typesof effects is particulate matter.

Past studies have demonstrated that higher levels of total suspended particulates lead to increasesin society's total expenditures on cleaning (both labor and cleaning supplies). In a study of the costs tohouseholds of elevated particulate levels conducted in the early 1980s, Mathtech found that the benefits ofreductions in particulates were significant. For each microgram per35

Lareau, Thomas J., and Douglas A. Rae, "Valuing WTP for Diesel Odor Reductions: An Application36

of Contingent Ranking Technique," Southern Economic Journal, January, 1989, pp. 728-742.

Nelson, Jon P., "Measuring Benefits of Environmental Improvements: Aircraft Noise and Hedonic37

Prices," Advances in Applied Microeconomics, Vol. 1, 1981, pp. 51-75.

Schulze, William D., et. al., Valuing Eastern Visibility: A Field Test of the Contingent Valuation38

Method, prepared for U.S. EPA, Office of Policy, Planning, and Evaluation, June 1991.

Rae, Douglas A., Benefits of Visual Air Quality in Cincinnati—Results of a Contingent Ranking39

Survey, draft report prepared by Charles River Associates for Electric Power Research Institute, 1984;


cubic meter reduction in particulate levels, they estimated a benefit of $0.31 per household. In the case ofthe particulate reductions under consideration at that time, summing these reductions across all affectedhouseholds yielded sizeable national benefits -- between $2 and $5 billion (1990 dollars). While thenumber of households affected by the proposed rule is smaller than the universe analyzed by Mathtech, thelocalized impacts of combustion facilities might be significant, particularly where there are large reductionsin particulate matter around waste burning cement kilns.

5.4.2.2 Aesthetic Damages

Reductions in emissions at waste combustion facilities or facility closures, perhaps due to increasedwaste minimization incentives, could have a variety of aesthetic benefits for nearby residents. These mightinclude reductions in odor, noise, or traffic, as well as changes in air quality that result in improvementsin regional visibility. A wide variety of studies suggest that the public values these types of environmentalimprovements. Because of the non-market nature of these types of disamenities, most studies rely onsurvey (contingent valuation) techniques to elicit willingness to pay for reductions in the disamenity. Forexample, one study surveyed individuals regarding willingness to pay for reductions in diesel odors. The36

study found a statistically significant annual willingness to pay of between $6 and $21 (1989 dollars) perhousehold for elimination of weekly "odor events." Likewise, researchers have demonstrated a significantproperty value impact from noise pollution. For example, a study of airport noise found an average priceloss of about $2,700 (1981 dollars) per home in areas exposed to peak noise levels. 37

While it is clear that aesthetic disamenities such as odor and noise can create economic damages,it is less clear whether the MACT standards would reduce or eliminate such disamenities. Odor problemsare likely to be specific to the wastes burned, and less directly associated with incremental reductions ofspecific pollutants. At facilities where waste burning is discontinued, there may be less frequent shipmentsto the facility, reducing traffic. Similarly, at incinerators that cease operation, there may be significantnoise reductions. Overall, however, it seems unlikely that major odor or noise reduction benefits wouldbe derived from the emissions reductions themselves.

Emissions of particulates may also have impacts on regional visibility. Previous studies havedemonstrated that the public is willing to pay for visibility improvements. These visibility improvementsmay enhance residential conditions or recreational experiences (e.g., long-range visibility inhiking/sightseeing areas). For example, an EPA-sponsored study used contingent valuation techniques tovalue a 2.4 mile visibility increase in urban areas. Respondents expressed an annual willingness to pay38

of between $18 and $39 (1990 dollars) per household. Earlier contingent valuation studies found evenlarger annual household willingness to pay estimates ($118 to $582, 1990 dollars) for greater changes invisual range. 39

Brookshire, D., et. al., Methods Development for Assessing Air Pollution Control Benefits, Vol. 2, preparedfor U.S. EPA, 1979; and Tolley, George, et. al., Establishing and Valuing the Effects of Improved Visibilityin Eastern United States, prepared for U.S. EPA, Office of Policy, Planning, and Evaluation, 1986.

Deposition of Air Pollutants to the Great Waters, First Report to Congress, prepared by EPA Office40

of Air Quality Planning and Standards, May 1994, p. 26.

Regulatory Impact Analysis of the Final Great Lakes Water Quality Guidance, prepared for U.S. EPA,41

January 5, 1995, p. 7-10.

Summary of Potential Economic Damages to the Resources of Lavaca Bay and Matagorda Bay, Texas42

as a Result of Mercury Contamination, prepared for the National Oceanic and Atmospheric Administration,May 1991.


The evidence from past studies suggests that the proposed rule might yield benefits in these areas.EPA is currently considering whether further analysis of these issues is warranted.

5.4.2.3 Recreational and Commercial Fishing Impacts

Releases from hazardous waste combustion facilities have the potential to adversely affectfisheries. The analysis of human health effects suggests that mercury poses the greatest concern, and couldlead to adverse exposures for anglers living around waste combustion facilities. If concern about mercurylevels leads to reductions in commercial or recreational fishing activities around waste combustionfacilities, society will experience a loss of economic welfare.

Past studies indicate that fish advisories for mercury occur frequently and can have substantialeconomic costs. For example, a study on mercury deposition to the Great Lakes demonstrates the linkbetween mercury deposition to surface water and fish consumption advisories. Both Lake Superior andLake Michigan as well as several rivers have experienced fish advisories for various species. The cost40

of these advisories in terms of lost recreational opportunities can be significant. For instance, a case studyof toxics pollution in the Saginaw River/Bay estimated annual recreational fishing losses of between$60,000 and $809,000 due to mercury and other toxic pollution. Another study of Lavaca Bay, Texas41

estimated lost recreational value of between $60,000 and $450,000 annually due to mercurycontamination. These site-specific figures represent relatively modest benefits; however, when we42

consider that mercury contamination from waste burning may affect multiple sites, the potential forsignificant aggregate benefits exists.

If reductions in mercury emissions from waste combustion facilities have the potential to eliminatefish advisories or allay public concerns in other ways that increase commercial or recreational fishing, theremay be benefits of this type for the proposed rule. EPA is continuing to investigate whether these typesof benefits are significant.


5.5 CONCLUSIONS

This chapter considers several different approaches to characterizing the benefits of the MACTstandards for hazardous waste combustors. The key findings are as follows:

! One method of gauging the benefits of the MACT standards is to consider the costper unit of emissions reduced. As would be expected, control of dioxin andmercury accounts for the greatest share of expenditures, and the marginal costs(expenditures per unit reduced) increase with more stringent controls. Across allHAPs, cost per ton of emissions reduced ($5,000 to $8,500 per ton) is somewhatabove, but generally comparable to other combustion rules for municipal andmedical waste combustors.

! Population risk estimates are significant for dioxin and mercury in the pre-regulatory baseline. EPA estimates that hazardous waste burning sourcesrepresent about nine percent of total anthropogenic dioxin emissions and aboutfour percent of total anthropogenic mercury emissions in the U.S. Humanexposure to dioxin may cause as many as 600 cancer cases each year in the UnitedStates. For mercury, health risks include severe neurological effects in theoffspring of exposed pregnant women and neurological effects in adults.

! Under the proposed MACT standard, reductions in dioxin and mercury emissionsfrom hazardous waste burning sources are significant. EPA expects that thesereductions, in conjunction with reductions in emissions from other dioxin andmercury-emitting sources, will help reduce dioxin and mercury levels over timein foods used for human consumption and therefore, reduce the likelihood ofadverse health effects, including cancer and neurological effects in adults, anddevelopmental abnormalities in children.

! EPA developed a screening analysis of ecological risks that considers modeledwatershed concentrations relative to risk-based ambient water quality criteria. Theanalysis suggests that water quality criteria currently may be exceeded in the mostvulnerable watersheds around waste combustion facilities, particularly cementkilns.

! Combustion of hazardous waste may also lead to economic damages that wouldbe reduced by the MACT standards. While investigation of economic benefits isnot yet complete, preliminary analysis reveals that hazardous waste combustionfacilities may reduce surrounding property values. If distance from a combustionfacility provides a reasonable proxy for the value of emissions reductions, thescreening analysis performed here indicates that benefits range from the tens tohundreds of millions of dollars per year. These estimates may reflect concernsover health risks and aesthetic disamenities and so cannot simply be added tobenefits in these areas.

Industrial Economics, Inc., "Seminar on Regulatory Impact Analyses and Regulatory Flexibility1

Analysis," prepared for the Office of Solid Waste, U.S. EPA, March 7, 1984; Habicht, F. Henry II, DeputyAdministrator, EPA. "Revised Guidelines for Implementing the Regulatory Flexibility Act," Office of theAdministrator, U.S. EPA, April 1992.


REGULATORY FLEXIBILITY ANALYSIS CHAPTER 6____________________________________________________________________________________

The Regulatory Flexibility Act (RFA) of 1980 requires Federal agencies to consider impacts on"small entities" throughout the regulatory process. Section 603 of the RFA calls for an initial screeninganalysis to be performed to determine whether small entities will be adversely affected by the regulation.If affected small entities are identified, regulatory alternatives must be considered to mitigate the potentialimpacts. Small entities as described in the Act are only those "businesses, organizations and governmentaljurisdictions subject to regulation." Although the RFA requires a regulatory flexibility analysis to beperformed only if a rule has a significant impact on more than 20 percent of the affected small entities, theEnvironmental Protection Agency (EPA) considers any impacts on any small entities to warrant furtherevaluation. As a result, we evaluate the impact of the proposed rule on all small entities.1

Evaluating the impact of the proposed maximum achievable control technology (MACT) standardson combustors involves first identifying which combustors can be classified as small businesses. We thenfurther evaluate the likelihood of negative impacts. Note that the screening analyses conducted in earlierchapters evaluated the impact of the proposed regulations at the combustion-unit level. This approachprovided the greatest insights into the decision to burn or not to burn hazardous wastes. However, theregulatory flexibility analysis focuses at the company level. This is because a large business that owns asmall combustion unit is not properly classified as a small business.

EPA, op. cit., "Appendix D: Developing Regulation-Specific Definitions of Small Entities."2

In certain instances, data on the parent company were either not available or contained sales and3

employment information for the headquarters location only rather than for the entire firm. In such cases,where headquarters information suggested a small business classification, we evaluated facility-levelinformation as well.


6.1 METHODOLOGY FOR IDENTIFYINGSMALL ENTITIES AFFECTED BY THE PROPOSED RULE

6.1.1 Statutory Definition

The RFA defines "small business" in the same manner as Section 3 of the Small Business Act. Thesmall business administration (SBA) defines "small" as a business that is independently-owned and notdominant in its field. Size thresholds, in terms of annual revenues or number of employees, vary bybusiness area as classified in the Standard Industrial Classification (SIC) and are drawn from 13 Code ofFederal Regulations (CFR) Part 121. Appendix F presents SBA thresholds for combustion facilities, aswell as data on actual sales and employee levels at these facilities. Other small entities include small not-for-profit organizations (e.g., private hospitals) and small governmental entities. However, government-2

owned combustion units affected by this proposed rule are owned by the federal government and wouldnot therefore be subject to regulatory flexibility considerations as a small entity.

6.1.2 Methodology for Evaluating Combustion Units for Regulatory Flexibility Analysis

To evaluate whether companies that own combustion units are small businesses, we developed alist of combustion units that included financial information at the facility and parent company levels. Thisinformation was originally compiled from Dun & Bradstreet and the American Business Director for EPA's"Resource Conservation and Recovery Act (RCRA) Expanded Public Participation and Revisions toCombustion Permitting Procedures" proposed rule in June 1994. We have updated the information wherepossible, including only combustion units on EPA's list of permitted combustion facilities from November1994.

Company data were then compared to statutory thresholds on employment and annual sales definedin 13 CFR Part 121 using the firm's primary SIC. Combustion units were classified as small businessesonly if the sales or number of employees at both the facility-level and the parent company level fell belowthe SBA threshold. In addition, we classified any firm for which we had no information on revenues and3

employment as a small business. The lack of information leads us to overstate the number of smallbusinesses.

See Jeffrey Smith, "Industrial Furnaces 1994," EI Digest, October 1994, pp. 17-25.4


6.2 NUMBER OF COMBUSTION FACILITIES DEFINEDAS SMALL BUSINESSES AND LIKELY IMPACTS OF COMPLIANCE WITH THE PROPOSED RULE

Given the capital intensity of cement production, commercial incineration, and many of theindustries (e.g., chemicals) that own and operate on-site incinerators, it is not surprising that few meet thedefinition of a small business. From a compiled list of more than 250 combustion units, only 13 wereclassified as owned by small businesses. As shown in Exhibits 6-1 and 6-2, five of these 13 wereconsidered small due to a lack of data. Of the remaining eight units, only three were owned by firms withannual revenues of less than $50 million per year, suggesting that financing new capital expenditures wouldbe less of a barrier than for smaller firms.

Although five units were classified as "small" because they did not employ enough people to meetthe SBA employment threshold, the companies that owned these units earn more than $50 million per yearin revenues. In three of these cases, annual revenues exceeded $100 million per year. Thus, despite thesmall number of employees, compliance with the proposed MACT standards is likely to pose a lowerfinancial burden.

Exhibit 6-2 lists the subset of combustion units classified as small businesses. Three commercialcombustors (Giant Cement, National Cement, and LWD, Inc.) burn sufficient waste quantities to remainprofitable under the most likely regulatory scenarios, although some may consolidate waste burning activityinto fewer units. The Featherlite unit does not show up on trade publication lists of active waste-burningkilns, suggesting that it may have already exited the market.4

Three on-site incinerators for which we have 1991 Biennial Report Survey (BRS) data on wastequantities burned (Aztec Catalyst, Cook Composites, and Parkens International) all burn less than 700 tonsof hazardous waste per year. Assuming they burn the average mix of liquids and solids for the on-siteincinerators we have data on, shipping the wastes to an off-site commercial facility would cost about $450per ton. This comprises between 0.1 and 0.2 percent of parent company revenues for the two facilities wehave tons burned data on, although impacts on plant-level operations would be higher. Given the patternof impacts on firms for which we have data, we do not expect significant impacts on the handful of firmsfor which we have no information on quantities burned or parent company revenues.

6.3 CONCLUSIONS

The proposed rule is unlikely to adversely affect small businesses for two important reasons. First,few combustion units are owned by businesses that meet the SBA definition as a small business.Specifically, available data allow us to identify only eight small entities out the more than two hundredEPA-listed combustion units burning hazardous waste. Furthermore, over one-half of those that areconsidered small have a relatively small number of employees, but have annual sales in excess of $50million per year.

Exhibit 6-1Summary of Combustion Units Subject to Regulatory Flexibility Assessment

Total

13Total Number of Units Considered "Small"5 Considered Small Due to Absence of Data

8 Classified as Small Based on Available Data3 Parent Revenues < $50 million/year2 Parent Revenues > $50m < $100m/yr.2 Parent Revenues > $100m < $250m/yr.1 Parent Revenues > $250m/yr.

Exhibit 6-2: Detail on "Small" Combustion Units

Parent-Level Small Business Screen

OverallP_ClassSourceP_salesP_nameF_ClassF_salesTypeBRS TonsEPA I.D.Commer.StateCityNameSize Class. Facility?

SNANA0AGGREGATE KILNTXD988040747YTXRangerFeatherliteSSABD/D&B$81,900,000GIANT GROUP LTDS6,492,000CEMENT KILNSCD003351699YSCHarleyvilleGiant CementSSD&B$81,900,000GIANT GROUP LTDS40,000,000CEMENT KILN41018.406PAD002389559YPABathKeystone Cement Co.SS$119,900,000NATIONAL CEMENT COMPANYNACEMENT KILNCAD982444887YCALebecNational Cement Co.SSABD$24,305,000AZTEC CATALYST CO.NA0INCINERATOR46.615OHD046202602NOHElyriaAztec Catalyst Resources/Dart Indst.SSABD/D&B$156,491,000COOK COMPOSITES & POLYMERSS10,302,000INCINERATOR697.335VAD055046049NVAChathamCook Composites & Polymers Co.SNANA0INCINERATOR371.58TXD008105959NTXHoustonParkens InternationalSNAS0INCINERATORLAD008213191NLAGeismarRubiconSSD&B$280,000,000SCHENECTADY INTERNATIONALNA0INCINERATOR8451.9NYD002070118NNYSchenectadySchenectady InternationalSNANA0INCINERATORWAD981765720YWABeverlyGrant County Waste Mgmt.SNAD&BS700,000INCINERATORTXD010791184YTXPasadenaHouston Chemical ServicesSNAD&BS1,146,000INCINERATORCAD983591660YCABoulevardLa Posta Recycling CenterSNANA0INCINERATOR25489.166KYD088438817YKYCalvert CityLWD, Inc.


Second, facilities most hurt by the rule tend to be those that burn very little waste and hence facevery high costs per ton burned. Our screening analyses and breakeven quantity analyses presented inChapter 3 demonstrate that the impact of the proposed rule is a function of tons burned rather than firmsize. Although we do not have data on the waste combusted at every small unit, those that have highcapacity utilizations will face relatively small cost increases per ton. Those that burn very little waste intheir existing units will close rather than comply with the proposed rule. Since their low quantities led tothe closure, it follows that their cost of off-site treatment will also be low.


ENVIRONMENTAL JUSTICE CHAPTER 7____________________________________________________________________________________

The U.S. EPA completed analyses that identified demographic characteristics of populations nearcement plants and commercial hazardous waste incinerators and compared them to the populations ofcounty and state. The analysis focuses on the spatial relationship of cement plants and incinerators tominority and low income populations. The study does not describe the actual health status of thesepopulations, and how their health might be affected by proximity to facilities.

7.1 METHODOLOGY

EPA used digital geographic and demographic data to develop population estimates around thecement plants and incinerators. EPA processed the Census geographic and demographic data along withthe incinerator and cement plant location using Geographic Information System (GIS) software. Theagency generated concentric circles or buffers for six radii (0.5,1,2,3,4,5 miles) from the site location andcomputed populations estimations for the six buffer areas. The finest possible data resolution was used -"block" or "block group" (as opposed to zip code or census tract level). EPA then compared thedemographics of the buffers to the demographics of the county and state that the facility was in.

EPA used the universe of 29 hazardous burning cement plants and a sample of 12 non-hazardousburning cement plants for a total of 41 cement plants from a universe of 113 cement plants. The summarystatistics presented for cement plants then uses the universe of hazardous waste burning plants and anextrapolation of non-hazardous burning plants from the sample of 12. EPA used a sample of 21commercial incinerators from a universe of 35. The complete methodology results of the analyses arefound in two reports filed in the docket titled, Race, Ethnicity, and Poverty Status of the Populations LivingNear Cement Plants in the United States and Race, Ethnicity, and Poverty Status of the Populations LivingNear Commercial Incinerators.

7.2 SUMMARY RESULTS

The Agency looked at whether minority percentages within a one mile radius are significantlydifferent than the minority percentages at the county for all hazardous burning cement plants and sampleof incinerators, the results are as follows:


! 27% of the universe of all cement plants (29 plants) and 37% of the sample ofincinerators (21 plants) have minority percentages within a one mile radius whichexceed the corresponding county minority percentages by more than fivepercentage points.

! 36% of the universe all cement plants (41 plants) and 44% of the sample ofincinerators have minority percentages within a one mile radius which fall belowthe corresponding county minority percentages by more than five percentagepoints.

! 38% of the universe of all cement plants (43 plants) and 20% of sample of theincinerators minority percentages within a one mile radius which fall within fivepercentage points (above or below) of the corresponding county minoritypercentages.

The Agency also examined whether poverty percentages within a one mile radius differsignificantly from the poverty percentages for the county as a whole. The results are as follows:

! 18% of the universe of all cement plants (20 plants) and 36% of the sample ofincinerators (21 plants) have poverty percentages at a one mile radius whichexceed the corresponding county poverty percentages by more than fivepercentage points.

! 22% of the universe of all cement plants (25 plants) and 37% of the sample ofincinerators (21 plants) have poverty percentages at a one mile radius which fallbelow the corresponding county poverty percentages by more than five percentagepoints.

! 60% of the universe of all cement plants (68 plants) and 28% of the sample ofincinerators (21 plants) have poverty percentages at a one mile radius which fallwithin five percentage points (above or below) of the corresponding countypoverty percentages.

7.3 LIMITATIONS

EPA excluded Puerto Rican facilities from the sample of cement plants analyzed because thecensus data required for the analysis is not readily available for Puerto Rico. In addition, the CensusBureau does not collect race information for Puerto Rico because the population is assumed to be ofHispanic origin. The results of both analyses may underestimate the minority populations near cementplants and commercial incinerators since these facilities were excluded.

The population estimated has important limitations. One limitation is the assumption that thepopulation is evenly distributed across a census polygon. Another limitation to the analysis is the resultof limitations inherent in the U.S. Census data. For example, the census data used is from 1990; therefore,there are issues with changes of populations characteristics over time. Also, the Census Bureau does notdetermine poverty status for all people (e.g. persons in college or in prison). Therefore, the poverty


statistics may underestimate the below poverty populations. Lastly, the accuracy of the facility locationand the quality of the location coordinate will also affect the number of people represented around afacility.

DRAFT: November 13, 1995

LIST OF ACRONYMS

ACFM Actual Cubic Feet per MinuteAPCD Air Pollution Control DeviceATF Above-the-FloorATTIC Alternative Technology Information CenterBDAT Best Demonstrated Available Technology BEQ Breakeven QuantityBIF Boiler or Industrial FurnaceBRS Biennial Report SurveyCEM Continuous Emissions MonitoringCERCLA Comprehensive Environmental Response, Compensation and Liability ActCETRED Combustion Emissions Technical Resources DocumentCIF Cost, Insurance and FreightCFR Code of Federal RegulationsCK Cement KilnCKD Cement Kiln DustCKRC Cement Kiln Recycling CoalitionCl Chlorine2

CO Carbon MonoxideCRF Capital Recovery FactorCWA Clean Water ActD/F Dioxin/FuranDOM Design, Operation, and MaintenanceDPRA DPRA, IncorporatedDRE Destruction and Removal EfficiencyEER Energy and Environmental Research CorporationEPA Environmental Protection AgencyESPs Electrostatic PrecipitatorsGDP Gross Domestic ProductGPM Gallons per MinuteHAP Hazardous Air PollutantHC HydrocarbonsHCl Hydrochloric AcidHg MercuryHSWA Hazardous and Solid Waste AmendmentsHWIR Hazardous Waste Identification RuleICR Information Collection RequestIWS Ionizing Wet ScrubbersLDR Land Disposal RestrictionsLVM Low Volatile MetalsLWA Lightweight AggregateLWAK Lightweight Aggregate KilnsMACT Maximum Achievable Control TechnologyNACR National Association of Chemical RecyclersNSPS New Source Performance StandardsO&M Operating and Maintenance


LIST OF ACRONYMS(continued)

OAQPS Office of Air Quality Planning and StandardsOMB Office of Management and BudgetOSW Office of Solid WastePCDD Polychlorinated Dibenzo-P-DioxinsPCDF Polychlorinated Dibenzo FuransPCI Pollution Control IndustriesPIC Products of Incomplete CombustionPM Particulate MatterPOTW Publicly Owned Treatment WorkPSPD Permits and State Programs DivisionRCRA Resource Conservation and Recovery ActRFA Regulatory Flexibility ActSBA Small Business AdministrationSQB Small Quantity BurnerSVM Semi-Volatile MetalsTEQ Dioxin/Furan Toxic EquivalentsTHC Total HydrocarbonsVISITT Vendor Information System for Innovative Treatment Technologies


REFERENCES

"1993 Outlook for Commercial Hazardous Waste Management Facilities: A Nationwide Perspective," TheHazardous Waste Consultant, March/April 1993.

"1994 Outlook for Commercial Hazardous Waste Management Facilities: A North American Perspective,"The Hazardous Waste Consultant, March/April 1994.

Bernstein, Leopold, "Financial Statement Analysis: Theory, Application and Interpretation," Howewood,IL: Irwin, pp. 640-652, 1983.

Black, Bob and Kitty Houghton, Industrial Economics, Inc. "Small Quantity Burners Research andObstacles Encountered," memorandum prepared for Lisa Harris, Regulatory Analysis Branch.March 17, 1995.

Boddy, Michelle, Energy and Environmental Resources Corporation, personal communication, March 27,1995.

Brigham, Eugene and Louis Gapenski, "Financial Management Theory and Practice," 6th Edition,The Dryden Press, p. 483, 1991.

Bureau of National Affairs. "Cement Industry `Enforceable Agreement' Would Replace Agency's Plan forKiln Dust," Environment Reporter, March 31, 1995, p. 2371.

Cement Kiln Recycling Coalition. "The Combustion Strategy and the CKRC," EI Digest, December 1993.

Clark, Wyman and Bruce Springsteen, Energy and Environmental Research Corporation. "MACT Costsfor New Sources," prepared for Frank Behan and Bob Holloway, U.S. Environmental ProtectionAgency, Office of Solid Waste, March 31, 1995.

"Commercial Hazardous Waste Management: Financial Performance and Outlook for the Future," annualreport from the July/August 1985, 1987, 1988, 1991, and 1993 issues of The Hazardous WasteConsultant.

Dempsey, Clyde and Timothy Oppelt. "Incineration of Hazardous Waste: A Critical Review Update," Air& Waste, Volume 43, January 1993, pp. 47-48.

Dickson, James and Jennifer Voltava, ICF Incorporated. "Information for Use in Categorizing FirmsOperating Hazardous Waste Incinerators and Boilers and Industrial Furnaces as `Small' Under SBAGuidelines," memorandum to Tracy Alfredson, U.S. Environmental Protection Agency, March 19,1994.

Energy and Environmental Research Corporation. Cost Estimates for APCD Requirements for ExistingFacilities to Meet Proposed MACT Standards for the Floor and Above the Floor Options forCement Kilns, Lightweight Aggregate Kilns, and Incinerators, prepared for U. S. EnvironmentalProtection Agency, Office of Solid Waste, April 1, 1995.


Energy and Environmental Research Corporation. "Summary of Penta Engineering Review ofEPA/OAQPS Cost Models for Cement Kiln Air Pollution Control Devices," prepared for FrankBehan, U.S. Environmental Protection Agency, Office of Solid Waste, April 5, 1995.

"Firing Up Discussions on a Clean Fuels Exemption," EI Digest, April 1995.

Fortuna, Richard. "The Combustion Strategy and the HWTC," EI Digest, December 1993.

Goebel, Chris, National Association of Chemical Recyclers, personal communication, August 11, 1994.

Goebel, Chris, National Association of Chemical Recyclers, personal communication with fuel blenders,including Brian Dawson (Brian Dawson and Associates); Bob Campbell (Pollution ControlIndustries); Brad Lamont (Romic Environmental Technologies); Scott Ellis (Cadence); GeorgeAnderson (Waste Research); February 14 and 22, 1995.

Graff, Robert and Thomas Walker. "Factors that Require, Encourage, or Promote Combustion ofHazardous Waste," memorandum to Walter Walsh, Office of Policy Analysis, U.S. EnvironmentalProtection Agency, prepared by Industrial Economics, November 11, 1993.

Habicht, F. Henry II, Deputy Administrator, U.S. Environmental Protection Agency. "Revised Guidelinesfor Implementing the Regulatory Flexibility Act," memorandum to Assistant Administrators,Regional Administrators, and General Counsel, April 1992.

Hanke, Jon. "Hazardous Waste Incineration 1993," El Digest, May 1993.

Hanke, Jon. "Hazardous Waste Incineration 1994," EI Digest, June 1994.

Hanke, Jon. "Hazardous Waste Incineration 1995," EI Digest, May 1995.

Hanke, Jon. "Hazardous Waste Landfills 1994," EI Digest, April 1994.

Hanke, Jon. "The Search For Continuous Emissions Monitors," EI Digest, March 1995.

Hoffman, Michael. "Waste Minimization Versus Segmentation," EI Digest, January 1994.

Holman, Hugh and D. Cotton Swindell. "Environmental Research: Chemical Waste Management andRollins Environmental Services," Alex Brown & Sons, Inc., April 5, 1993

ICF Incorporated. 1990 Survey of Selected Firms in the Hazardous Waste Management Industry , July1992, prepared for the U.S. Environmental Protection Agency, Office of Policy Analysis.

Industrial Economics, Inc. "Seminar on Regulatory Impact Analyses and Regulatory Flexibility Analysis,"prepared for the Office of Solid Waste, U.S. Environmental Protection Agency, March 7, 1984.

Industrial Economics, Inc. Analysis of Compliance Costs Associated with Proposed Dioxin Rule forHazardous Waste Combustion Facilities, Draft, prepared for the U.S. Environmental ProtectionAgency, Office of Solid Waste, August 1994.


Koplow, Doug, Bob Black and Tom Walker, Industrial Economics, Inc. "Key Regulatory InitiativesImpacting Hazardous Waste Combustion," memorandum prepared for Lisa Harris and SaraRasmussen, U.S. Environmental Protection Agency, November 22, 1994.

Koplow, Doug, Daphne McMurrer, Kitty Houghton, Anne Petersen, Bob Black and Tom Walker,Industrial Economics, Inc. "Preliminary Analysis of the Impact of Source Reduction, Recycling,and Increased Compliance Costs on Hazardous Waste Combustion," memorandum prepared forLisa Harris and Sara Rasmussen, U.S. Environmental Protection Agency, September 20, 1994.

Kryder, Greg and Wyman Clark. "Revised MACT Compliance Costs," memorandum prepared by Energyand Environmental Research Corporation for Larry Denyer, U.S. Environmental ProtectionAgency, April 24, 1995.

Kryder, Greg and Bruce Springsteen, Energy and Environmental Resources, memorandum to DougKoplow, Industrial Economics, June 27, 1995.

Kyder, Greg, et al. "Shutdown Time Estimates," memorandum prepared by Energy and EnvironmentalResearch Corporation for Bob Holloway, U.S. Environmental Protection Agency, May 5, 1995.

McMurrer, Daphne, Robert Black and Thomas Walker. "Memorandum: The Processing and Use of WasteFuels," prepared by Industrial Economics for Lisa Harris, U.S. Environmental Protection Agency,Office of Solid Waste, December 13, 1994.

Mulle, Thomas. "Cement and Aggregates," Value Line, January 20, 1995, p. 891.

National Association of Chemical Recyclers. NACR Waste Processing Survey, August 1994, question 1.

Nef, Wayne. "Rollins Environmental Services," Value Line, June 24, 1994, p. 352.

Nef, Wayne. "Rollins Environmental Services," Value Line, March 24, 1995, p. 350.

Perket, Cary and Jon Hanke. "The Roots of Overcapacity in the Incineration Sector," EI Digest, February1995.

Portland Cement Association. U.S. Cement Industry Fact Sheet Twelfth Edition, June 1994, p. 17.

Randolph, Karen. U.S. Environmental Protection Agency, personal communication, June 13, 1995.

"RCRA Expanded Public Participation and Revisions to Combustion Permitting Procedures; ProposedRule," Federal Register, June 2, 1994, pp. 28680 - 28711.

Retallick, Phil. Rollins Environmental Services, personal communication, September 13, 1994.

Seidel, Christine. "Another Look at the EPA's 1991 Biennial Report," EI Digest, March 1995.

Seidel, Christine. "Hazardous Waste Treatment and Disposal Summary 1995," EI Digest, January 1995,p. 6.


Seidel, Christine. "Mobile Thermal Treatment 1994," El Digest, December 1994.

Smith, Jeffrey. "Fuel Blenders 1994," El Digest, September 1994.

Smith, Jeffrey. "Industrial Furnaces 1993," EI Digest, September 1993.

Smith, Jeffrey. "Industrial Furnaces 1994," EI Digest, October 1994.

Strauss, Jerome. "Memorandum: Evaluation of Alternative Treatment Technologies for Wastes that AreCurrently Being Sent to Combustion Facilities (Revised-Draft)," prepared for U.S. EnvironmentalProtection Agency, March 30, 1995.

U.S. Environmental Protection Agency. 1991 Biennial Report, Draft Data Summaries, September 29,1994.

U.S. Environmental Protection Agency. Combustion Emissions Technical Document (CETRED), May,1994.

U.S. Environmental Protection Agency, Estimating Exposure to Dioxin-Like Compounds, Volume I, June1994.

U.S. Environmental Protection Agency, Estimating Exposure to Dioxin-Like Compounds, Volume II , June1994.

U.S. Environmental Protection Agency, Health Assessment Document for 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds, Volume II, June 1994.

U.S. Environmental Protection Agency, Health Assessment Document for 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds, Volume III, August 1994.

U.S. Environmental Protection Agency, Mercury Study Report to Congress Volume II: Inventory ofAnthropogenic Mercury Emissions in the United States, December 1994.

U.S. Environmental Protection Agency, Mercury Study Report to Congress Volume III: An Assessmentof Exposure from Anthropogenic Mercury Emissions in the United States , December 1994.

U.S. Environmental Protection Agency, Mercury Study Report to Congress Volume VI: Characterizationof Human Health and Wildlife Risks from Anthropogenic Mercury Emissions in the United States ,December 1994.

U.S. Environmental Protection Agency, National Study of Chemical Residues in Fish, Office of Scienceand Technology, September 1992.

U.S. Environmental Protection Agency. Report to Congress on Cement Kiln Dust, 1993.

U.S. Environmental Protection Agency. Setting Priorities for Hazardous Waste Minimization, July 1994,p. 2-12.


U.S. Environmental Protection Agency. Technical Background Document: Economic Impact Analysis forthe Proposed RCRA Expanded Public Participation and Revisions to Combustion PermittingProcedures Rule, Draft, May 16, 1994.

U.S. Environmental Protection Agency. Total Mass Emissions from a Hazardous Waste Incinerator,prepared by Midwest Research Institute for Accurex Corporation, June 12, 1987, p. 4-6.

U.S. Environmental Protection Agency, Office of Solid Waste. Background Document for CapacityAnalysis for Newly Listed and Hazardous Debris to Support 40 CFR 268 Land DisposalRestrictions (Final Rule), Volume 1: Capacity Analysis Methodology and Results , June 1992.

U.S. Environmental Protection Agency, Office of Solid Waste. Background Document for CapacityAnalysis for Land Disposal Restrictions Phase II - Universal Treatment Standards, and TreatmentStandards for Organic Toxicity Characteristic Wastes and Other Newly Listed Wastes (Final Rule)Volume I: Capacity Analysis Methodology and Results, August 1994.

U.S. Environmental Protection Agency, Office of Solid Waste and Emergency Response, NationalBiennial RCRA Hazardous Waste Report (Based on 1989 Data) , February 1993.

U.S. Environmental Protection Agency, Permits and State Programs Division. "Hazardous Waste Boilersand Industrial Furnaces (BIFs) with Either a Metals Reclamation Exemption or Small QuantityBurner Exemption," February 23, 1994.

regulatory impact assessment for proposed …regulatory impact assessment for proposed hazardous...

Documents